Skip to main content

Full text of "Illuminating engineering"

See other formats 


Digitized  by  the  Internet  Archive 

in  2011  with  funding  from 

University  of  Toronto 


http://www.archive.org/details/illuminatingengi08illu 


Transactions 


of  THE 


Illuminating  Engineering 
Society 


vol.  VIII 
JANUARY-  DECEMBER 

1913 


Subject  Index  and  Index  to  Authors 


ILLUMINATING  ENGINEERING  SOCIETY 

29  WEST  THIRTY- NINTH  STREET 
NEW   YORK 


/ 
loo 

M  33 


"><? 


SUBJECT  INDEX. 


*  Pages  following  the  numbers  given  should  be  consulted  in  referring 
to  a  topic  or  subject. 

Absorption  of  light :  page 

Aluminum  finished  reflectors   281,  290 

Prismatic  and  opal   glassware 452 

Reflectors  past  and  present 107 

Accidents : 

Caused  by  poor  lighting 139,  474 

Attributed  to  darkness  and  insufficient  light 134 

Prevention  by  proper  illumination  of  streets 91 

Acetylene : 

Lighting  status    333 

Spectrum    465 

Acuity.     (See  visual  acuity.) 

Agricultural  lighting,  low-voltage  lamps  and  systems 327 

Aluminum  reflectors    269,   271,  282 

Absorption   of   light    281,  290 

Altar  lighting   617 

Alternating  current  low  frequency  for  flame  carbon. arc  lamps 328 

Antique  and  curio  shop  lighting 524 

Architect : 

Need    of    co-operation    between    decorator    and    illuminating 

engineer   153 

And  illuminating  engineering   344 

Arc  lamps:     (See  also  flame  carbon  arc.) 

Device  for  converting  enclosed  alternating  and  direct  current 

carbon  arc  lamps  into  flame  arc  lamps 327 

Enclosed  carbon,  operation  characteristics  compared  with  flame 

carbon  arc  lamps  166 

Luminous   for  street  lighting 88 

Magnetic  blow    165 

Magnetite  arc,  increased  use 327 

Modification  of  light  from  carbon  arcs  to  approximate  daylight  352 

Non-magnetic  arcs    165 

Peculiar  phenomenon  of   positive  -crater 328 

Stability  of  arc  influenced  by  wind,  magnetic  fields,  etc 328 

■  Status  of  open  and  enclosed  series  and  multiple  arc  lamps....  327 

Street  lighting  installation  of  New  York 202 

Theory  of  luminosity    407 

Titanium  arc  characteristics    407 

Ultra-violet  light  radiation    165 

_ Variations  in  length  of  arc  with  voltage  changes 165 

vs.  tungsten  lamps  in  department  store  lighting 17 


IV  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 

Automobile  show  lighting  (gas) 265 

Barber  shop,  lighting   508 

Bedroom  lighting 259 

Book  store,  lighting   524 

Brightness:     (See  also  intrinsic  brilliancy.) 

A  new  instrument  for  observing  intrinsic  brilliancy 341 

Flame  arc  lamps   483 

Of  tungsten  lamp   filaments 353 

Of  various  colored  surfaces  exposed  to  different  illuminants . .  70 

Of   the  sky    233 

Test   for   discrimination    43 

Cadmium-vapor  tube  lamp,  efficiency 329 

Calculation : 

A  method  for  determining  illumination  at  any  point  on  a  flat 

surface    355 

Candle-power:     (See  also  Deterioration.) 

Cause  of  falling  off  in  mantles 332 

Deterioration  in  gas  lighting  units  attributed  to  mantle  burner 

and  glassware   332 

Of  high  pressure  gas  lamps 331 

Low  pressure  high  efficiency  gas  lamps  for  exterior  lighting . .  330 

Tests  of  tungsten  lamps 550 

Variation    in    electric    incandescent    lamps    produced    by    bulb 

frosting    326 

Variation  in  electric  incandescent  lamps  with  voltage  change  595 

Candy  shop,  lighting    520 

Carbons,  arc.     (See  electrodes.) 
Carbon  electric  incandescent  lamps : 

Change  in  candle-power  with  variation  in  voltage 595 

Damage  in  show-case  lighting 2>7 

-, —  Method  of  testing  for  spots  in  filaments 326 

Percentage  of  total  incandescent  lamp  sales  for  three  months 

of    1913    656 

Relative  number  of  carbon,  gem,  tantalum  electric  incandescent 

lamps  sold  in  1907,  1908,  1909,  1910,  191 1,  1912 323 

versus  tungsten  lamps  for  seeing 146 

Carcel  lamp,   color  of   light 316 

Car  lighting.     (See  railway  car  lighting.) 

Carpet  store   508 

Cataract  of  the  eye  caused  by  injurious  light 132 

Chancel  lighting    617 

Chromatic  aberration  of  the  eye 298 

Church  lighting : 

Altar    617' 

Apparatus  and  fixtures   618 

Average  intensity  of  illumination 624 


SUBJECT   INDEX     '  •  V 

Church  lighting  {continued) : 

Bulletin  board  lighting   622 

Chancel   lighting    617 

Deck  system    614,  625,  627,  628 

Exterior  illumination   618 

General  requirements    617 

Use  of  indirect  lighting 615,  616,  623,  626 

Color : 

Artificial  daylight  booth   for   matching  colors 34,     38 

Comparative  diffusiveness  of  lights  of   different  colo; 232 

Light  for  theatrical  effects 35 

Loss  of  efficiency  of  the  eye  caused  by  lights  of  different  color  256 

—  In  lighting  installations   663 

—  Limen  test,  for  color  sensitivity  of  the  eye 41 

—  The  measurement  of  lights  of  different  color 302 

Perception  by  the  eye 343 

—  Psychic  value   of   illumination 357 

Visibility  of  railway  signal  lights  of  various  colors 299 

Colorimeter,  use  and  difference  from  spectro-photometer 71 

Colored  surroundings,  influence  on  color  of  useful  light 62 

Coefficients,   reflection    68 

Costs : 

Artificial  daylight  for  interior  illumination 253 

Department  store  lighting   ^2 

Industrial   lighting    476,   486,  672 

Railway  car  lighting    602 

Small  home  lighting  672 

Cove   lighting    615 

Deficiencies    114 

Daylight : 

An   approximation   of   an   interior   distribution   with   artificial 

light  sources    244 

An  example  of  a  copy  of  daylight  with  artificial  light  sources..  234 

Characteristics    233,  238 

Compared  with  artificial  for  interior  illumination 134 

Compared  with  indirect  lighting 626 

Cost  of   artificial  daylight 253 

Desirable  qualities  found  in  window  lighting  by  daylight 252 

Diffusion  of  daylight  compared  with  artificial  light  sources...  232 

Direction  and  distribution  of  daylight  in  interiors 113 

Efficiency  of  tungsten  lamp  approximating  the  color  of  day- 

light      326 

The  ideal  illuminant   404 

Intensities  compared  with  interior  illumination  intensities....   112 

' Intrinsic  brilliancy  of  the  sky 233 


vi  TRANS.    I.  K.  S.      VOL.  VIII 

PAGE 
Daylight  [continued) : 

Modification  of  light  from  electric  arc  lamps  to  approximate 

daylight    352 

— —  Simulation   for  interior  illumination 124 

Definitions.      (See  proposed  definitions,  page  6,  part   1,  June    (1913) 
issue.) 

Delicacies  shop,  lighting  528 

Department  store  lighting: 

—  Artificial  daylight  for  color  matching 34 

Avoidance  of  high  intrinsic  brilliancy 19 

Cleaning  lighting  equipment    20 

-. Color  of  light  required 19 

Costs  of  cleaning  and  maintenance 32 

Deterioration    37 

Fixtures    (direct)    21 

Important  considerations  involved    19 

Lighting  systems    17 

McCreery  &  Co.,   Pittsburgh,  store 17 

Requirements   19,  33,  36 

Rug  racks,  lighting   22 

Show  cases    30 

Show  windows  (see  subject :  Window  lighting) 28 

Deterioration : 

— i —  Illumination  in  a  department  store 37 

—  Of  light  from  aluminum  finished  reflectors 281,  290 

-  In  gas  lighting  units  attributed  to  mantle  burner  and  glass- 
ware      332 

Cause  of  candle-power  deterioration  of  mantles 332 

Diffuse  reflection : 

Explained  and  illustrated   270 

Diffusion : 

Its  importance  in  illumination 123,   125,  661 

Daylight  compared  with  artificial  light  sources 232 

Dining   room   lighting    259 

Direct  lighting : 

Compared  with  indirect  and  semi-indirect 116 

Compared  with  indirect  lighting  in  a  department  store 18 

Abuses  in  application    114 

Comparative  illuminations  produced  from  sources  of  the  same 

flux  by  direct,  indirect  and  semi-indirect  methods 118 

In  churches   .- 614 

Efficiency : 

Basis   for  comparing   reflectors 280 

Cadmium-vapor  tube  lamp  329, 

Daylight  approximation  with  use  of  tungsten  lamp 326 


SUBJECT    INDEX  Vli 

PAGE 
Efficiency  {continued): 

Eye  after  a  period  of  work 45 

Eye  under  lights  of  various  colors 145 

Flame  arc  lamps    328,  485 

Good  lighting  as  an  essential,  in  factories 137 

Loss  of  efficiency  of  the  eye  under  lights  of  different  color...  256 

Low  pressure  high  efficiency  gas  lamps  for  exterior  lighting..  330 

Of   mercury-vapor   lamp   combined   with   tungsten   lamp   light 

to  give  white  light 328 

Of  Moore  dioxid  lamp 376 

Of  Neon  tube  lamp 376 

Of  nitrogen,  filled  tungsten  lamp 325 

Of  various  types  of  reflectors 289 

Tests  of  the  eye  under  different  systems  of  lighting 41 

Of  various  tungsten  lamps 324 

Quartz  tube  lamp  for  street  lighting 329 

Electrodes : 

Flame  electrodes.     (See  Flame  carbon  arc  lamps.) 

Enclosed  carbon  operation  characteristics  compared  with  flame 

carbon  arc  lamps   166 

Peculiar  phenomenon  of  positive  crater 328 

Titanium  arc  characteristics    407 

■  Ultra-violet  light  radiation    '. 165 

Electron  theory,   mercury-vapor   lamps 77,  86 

Eye:     (See  also  Visual  acuity.) 

■  Angle    of    direction   of    light,    its    effect    on    the    efficiency   of 

the  eye    277 

Bottle  makers'  cataract  and  an  explanation  of  its  cause 132 

Causes  of   discomfort    54 

Chromatic  aberration    298 

Discomfort,  indications,  caused  by  illumination 53 

Discomfort  and  influence  caused  by  direction  of  light 255 

Effect  upon  the  eye  of  ultra-violet  light  emitted  by  commer- 

cial light  sources   343 

Efficiency    256 

Efficiency  under  lights  of  various  color 145 

— •  Efficiency  under  several  systems  of  illumination 256,  257 

Elimination  of  glare — a  cause  of  .ocular  discomfort no 

r  Eyestrain ;  Its  nature  and  cause 141 

—  Fatigue     and     eyestrain     caused     by     watching     motion     pic- 
tures     188,   190,  191 

Influence  of  ultra-violet  light 132,  138 

Loss  of  efficiency  after  a  period  of  work 45,  51 

Loss  of  efficiency  under  lights  of  different  color 256 

Perception  of  color    343 


V1I1  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 
Eye  {continued) : 
Research  to  ascertain  effect  of  luminous  radiation  upon  the  eye  342 

Tests  for  brightness  discrimination 43 

Tests    for   color   sensitivity 42 

-  Tests  for  efficiency  and  discomfort  under  different  systems  of 

illumination    40,     58 

Tests  for  visual  acuity 43 

-  Vision  as  influenced  by  the  brightness  of  surroundings 

292,  297,  298,  299 

Walls :  An  element  affecting  ocular  comfort 108 

Eyestrain : 

And  illumination  in  industrial  buildings 277 

And   illumination    128,  130 

And  glare  from  tungsten  lamps  in  railway  cars 608,  609 

Exhibition  hall  lighting: 

Semi-indirect    (gas)    263 

Factory  lighting:     (See  also  Industrial  lighting.) 

Progress  and  status   486 

Costs   476,  486,  672 

Average  illumination  intensity    658 

Firefly : 

■ Intrinsic   brilliancy    335 

Fixtures : 

A  novel  fixture  in  a  state  building 351 

■  And  apparatus   for  church  lighting 618 

Department  store    (direct)    21 

Distinctive  store  lighting 522 

Hospital    489 

Poor    illumination    of    interiors    attributed    to    fixture    manu- 
facturers        154 

Railway  car  214 

Semi-indirect    264 

Substitution   of  marble   for   glassware 352 

■  Window  lighting   564 

Flame  carbon  arc  lamps  : 
— —  Applications    327 

Brilliancy,  intrinsic 483 

Candle-power  variation  with  voltage  and  current  variations..   171 

Comparative  light  giving  efficiency  of  yellow  and  white  carbons  175 

Candle-power   and   operation   characteristics 177,  328 

Color  of  light  depends  upon  current  density 164 

Composition,  color  of  light,  and  operation  characteristics 168 

Efficiency  of  yellow  and  white  carbons  for  industrial  lighting  484 

Factors    which   determine   light X73 


SUBJECT   INDEX  ix 

PAGE 
Flame  arc  lamps  {continued): 

Improvement  in  carbons    327 

Life   of   carbons    327 

Scheme  for  determining  amount  of  illuminants  in  an  electrode  170 

Comparative   performance   on    direct   and   alternating   current 

circuits    174 

Current  densities  of  electrodes  and  terminals 167 

Efficiency   328,  485 

Device  for  converting  enclosed  alternating  and  direct  current 

lamps  into  a  flame  arc  lamp 327 

Displacing  open  and  enclosed  arc  lamps 327 

For  low  frequency  circuits 328 

Controversy   with   high   pressure    gas   lighting   people   over   a 

street  installation  in  England 349 

Heights  for  street  installations 485 

Improvements  in  construction    327 

Magnetic  lamps   165 

Light  independent  of  polarity 328 

Maintenance    176 

Operation  characteristics   163,  166 

Peculiar  phenomenon  of  positive  crater 328 

Glassware    .' 175 

Globes  for  prevention  of  soot 172 

Slagging    171 

Stability  in  flame  carbon  arc  lamps 165,  328 

Advantages    for   street   lighting 175 

Street  lighting  in  Chicago 485 

Street  lighting  in  New  York 202 

Variations  in  length  of  arc  with  voltage  changes 165 

Theory  of  luminosity    407 

Ultra-violet  light  radiation    165 

Focusing  tungsten  lamp,  construction  and  efficiency 326 

Focusing   type    reflectors    274 

Gas  lamps : 

Cause  of  falling  off  of  candle-power 332 

Characteristics  of  a  high  power  single  mantle  inverted  lamp..  331 

Candle-power  deterioration  in  gas  lighting  units  attributed  to 

mantle  burner  and  glassware 332 

High  pressure.     (See  High  pressure  gas  lighting.) 

Modification  of  light  to  approximate  daylight 352 

New  types  of  units    330 

Number  used  in  street  illumination  in  New  York 202 

Artificial  silk  mantle    332 

Theory  of  luminosity  of  Welsbach  mantle 407 


X  TRANS.    I.  K.  S.      VOL.  VIII 

PAGE 

Gas  lighting:     (See  also  gas  lamps.) 

Acetylene    333 

Cause  of  falling  off  of  candle-power 332 

Golf  courts   347 

High  pressure  lamps,  difficulties  and  improvements 331 

Ignition,  elements  used   332 

Low  pressure  high  efficiency  lamps  for  exterior  lighting 330 

Candle-power  and  efficiency    331 

Natural  gas,  a  new  method  of  utilization 333 

New  types  of  units . , 330 

Gas  pressure  in  several  cities  of  United  States 5 13 

Semi-indirect  in  an  exhibition  hall 263 

In  small  stores    499 

In  a  Sunday-school  room 439 

Tennis  courts   347 

Gem  lamps : 

Relative  number  of  carbon,  gem,  tantalum  electric  incandes- 

cent lamps  sold  in  1907,  1908,  1909,  1910,  191 1,  1912 323 

Percentage  of  total  incandescent  lamp  sales  for  three  months 

of   1913    656 

Glare : 

Cause  of  drowsiness  during  church  services 624 

Elimination  by  location  of  light  sources no 

Influence  on  seeing   104,  109 

From   lamps    608 

Means  of  elimination  no 

Its  nature  and  character , 667 

Patented  bulb  for  reducing  glare  of  electric  lamp  filament....  327 

From  tungsten  lamps  in  railway  cars 608,  609 

Glassware : 

Characteristics    (distribution,   absorption,   color,  transmission) 

of  prismatic  and  various  enclosing  glassware 447 

Distribution    452 

Effect  upon  life  of  tungsten  lamps 456 

For  flame  carbon  arc  lamps 175 

Globes   for  the  prevention  of   soot  accumulation  from   flame 

carbon  arc  lamps 172 

Substitution  of  marble  for  glassware  on  lighting  fixtures....  352 

Golf   courts,   lighting    '. 347 

Great  white  way  lighting  90 

Grocery   store,  lighting    , 521 

Hefner  lamp  as  a  standard  in  photometry 412,  422' 

Heterochromatic  photometry,  a  practical  solution 302 


SUBJECT   INDEX  xi 

PAGE 
High  pressure  gas  lighting: 

Candle-power  of   lamps    33 1 

Controversy  with  electric  people  over  installation  in  England . .  349 

Difficulties  and  improvements  in  lamps 331 

Progress  and  status  of  street  lighting 350 

Street  lighting  status,  United  States  and  abroad 331 

Home  lighting : 

Advantages  and  disadvantages  of  home  experiments 230 

Experiments  in  illumination  from  large  area  light  sources....  229 

Influence    of    light    and    dark    walls    upon    the    illumination 

required    151 

Location  of   light  sources 126 

Requirements    151 

Simulation  of  daylight  for  interior  illumination 124 

Hospital  lighting: 

Corridors    498 

Operating   rooms    488,  498 

Operating  table  lamp  493 

Wards    494,  498 

Zeiss  system,  illumination  intensities ■ 497 

Illuminating  engineer : 

His  work  and  qualifications 100,   169,  669 

Need  of  co-operation  with  architect 153 

Illuminating  engineering : 

As  a  science  and  an  art 100 

And  architects    345 

Use  of   photography    554 

Laboratory  and  equipment  of  General  Electric  Co 379 

Illuminating  Engineering  Society: 

Classification  of  papers  6 

Committees  activities  ( 1912)    8 

Committee   reports  and   activities 689 

Finances    683 

Functions    3 

General    Secretary's    report    .' 6,  683 

History  and  growth   1 

Illumination  primer  "Light:  Its  Use  and  Misuse" 675 

Membership   classification    14,  684 

Its  unique  position  among  societies 676 

Its  relation  to  public  lighting  companies 99 

Report  of  General  Secretary 6 

Sections    687 

Transactions    688 


xii  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 
Illuminants :     (See  also  lamps  according  to  name.) 

Daylight,  the  ideal  illuminant 4°4 

Determination  of  radiation  efficiency 404 

Radiation  efficiency 404 

Ultra-violet  radiation  from  artificial  light  sources 132 

Use  of  spectrum  in  determining  value 407 

Illumination:     (See  also  lighting.) 

Angle  of  direction  of  light  for  efficiency  of  the  eye 277 

Average   intensities   of   store,   street,    show   window,    factory, 

office,  residence,  railway  cars 658 

Average  intensity  in  churches 624 

Comparative  illuminations  produced  from  sources  of  the  same 

flux  by  direct,  indirect  and  semi-indirect  methods 118 

Concerning  data  on   competitive  illuminants 514 

Contrast  in  illumination;  its  nature  and  importance 666 

Department  store  intensities   31 

Direction  and  distribution  of  daylight  in  interiors 113 

Elimination  of  streaky  effect  from  metal  reflectors 286 

And  eyestrain  128,  130,  277 

Inadequacy  of  measurements  in  the  horizontal  plane 158-160 

Increased  industrial  production  with  good  illumination 473,  485 

Increases  in  intensities  in  recent  years 655 

And  industrial  accidents    474 

Influence  of  colored  surroundings  on  the  color  of  the  useful 

light    61,  107 

Influence   of    decoration  and   contrast   in  the   illumination   of 

interiors .■ 108 

Influence  of  light  between  60°  and  horizontal  upon  sight 291 

Intensities.     (See  Illumination  intensities.) 

Intensity  required  for  street  lighting 89 

Interiors,  various  phases  and  problems 99 

Legislative  requirements  in  Holland,  England,  New  York  and 

Wisconsin    353 

A  method  for  determining  illumination  at  any  point  on  a  flat 

surface   355 

A  method  for  plotting  distribution  curves , 355 

Its  principal  use 101 

Progress  during  the  year  ... 323 

Psychology  of  intensities   284 

Relative  intensity  of  sunlight  and  artificial  illumination 301 

Reflectors.     (See  Reflectors.) 

Tests  for  the  efficiency  of  the  eye  under  different  systems  of 

lighting    40 

Typical  street  lighting  intensities 658 


SUBJECT   INDEX  xiii 

T  J  ,  PAGE 

Incandescent  electric  lamps:     (See  also  lamps  by  name.) 

Average   candle-power    of   all    incandescent   lamps   sold    1906- 

«W3    657 

Cooling  effect  of  leading-in  wires  upon  lamps  of  the  street 

series  type   3g5 

Helical  filaments   537 

Method  of  testing  for  spots  in  incandescent  lamps  filaments...  326 

Relative    number    of    carbon,    gem,    tantalum    and    tungsten 

electric  incandescent  lamps  sold  in   1907,   1908,   1909,    1910, 
191 1,    1912    32, 

A  new  method  of  determining  the  true  temperature  of  fila- 

ments          *,->! 

334 

Patented  bulb  for  reducing  glare  of  filament 327 

Recent  improvements  in  manufacture  534 

Smashing  point    S40)  546,  547 

Sign  lamps  ,24 

Indirect  lighting: 

Comparative  illumination  produced  from  sources  of  the  same 

flux  by  direct,  indirect  and  semi-indirect  methods  118 

Compared  with  daylight   626 

Compared  with  direct  and  semi-indirect   ' 116 

Compared  with  indirect  in  a  department  store 18 

Fixtures  for  railway  car  lighting  348 

Hospital  fixture   492 

In  churches   615,  616,  623,  626 

Its  influence  upon  the  design  of  lighting  installations 114 

Industrial  lighting: 

Accidents  caused  by  poor  illumination  and  darkness ..  134,  139,  474 

Costs    476,  486,  672 

Good  lighting  as  an  essential  to  efficiency 137 

In  factories    486 

Influence  of  good  lighting  on  production 287 

Legislative  requirements  in  Holland,  England,  New  York  and 

Wisconsin   353 

Mercury-vapor    lamps    479 

Metal  reflectors    268 

Progress 345,  486 

Percentage  increase  in  output  with  good  illumination 473,  485 

Standards  compared  with  store  illumination   283 

Total  annual  cost    487 

Intensities  of  illumination : 

Average  in  churches    624 

Department  store  illumination  32 

Factory    lighting 658 

Motion  picture  screens    195 

Increases  in  recent  years   655 


XIV  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 

Intensity  of  illumination  {continued): 

Industrial  lighting  658 

Office  lighting  658 

Railroad  car  lighting    223,  658 

Residence  lighting  658 

Show  window  lighting  658 

Store  lighting   658 

Street   lighting    658 

Intensive  type  reflector   274 

Interior  illumination : 

Requirements 101 

Various  phases  discussed  and  illustrated   99 

The  window  as  a  light  source 241 

Intrinsic  brilliancy : 

Of  firefly    335 

Flame  carbon  arc  lamps  483 

Of  the  sky  233 

Tungsten  lamp  filament 233,  353,  483 

Welsbach  mantle   233 

Jewelry  store  lighting    508,  517 

Ladies'  wear  shop  lighting. 525 

Lamps.     (See  lamps  according  to  name,  arc,  incandescent,  gas,  etc.) 
Kerosene  lamp :    A  reason  for  its  present  day  use  132 

Miniature  tungsten  lamps,  construction  details  325 

Carbon    arc    162 

Relative  number  of  carbon  gem,  tantalum,  and  tungsten  elec- 

tric   incandescent    lamps    sold    in    1907,    1908,    1909,    1910, 
1911,   1912    323 

Lamp-posts : 

New  York  street  lighting 200 

Ornamental   for  street  lighting 92 

Legislation,  lighting   347 

Industrial  illumination  requirements  in  Holland,  England,  New 

York,  and  Wisconsin  353 

Law  of  reflection    270 

Library  lighting : 

Investigation   conducted   by   British    Illuminating   Engineering 

Society 344 

Light : 

"Cold"  light   334 

Color  best  adapted  to  seeing 143 

Color  of  light  required  for  motion  picture  projectors 184 


SUBJECT   INDEX  XV 

PAGE 
Light  {continued): 

Direction  of  light  in  interiors 659 

Discomfort  and  influence  on  the  eye  caused  by  direction  of 

light    255 

Explanation  and  theory    131 

Kerosene  light  for  seeing 132 

Perception  of  lights  of  short  duration 342 

Poor  illumination  attributable  to  fixture  manufacturers 154 

Psychic  value  of  light,  shape,  form  and  color 357 

Radiation  of  carbon  incandescent  lamps  and  tungsten  lamps 

compared  for  seeing    146 

Some  theoretical  considerations  of  production 400 

Various  colors  and  their  influence  upon  vision 145 

Yellow  versus  white  light  for  seeing 145 

Lighting:     (See  also  indirect  and  semi-indirect  lighting;  illumination.) 

Abuses  in  application  of  direct  lighting 114 

Agricultural,  low  voltage  lamp  and  system 327 

Angle  of  direction  of  light  for  efficiency  of  the  eye 277 

Approximation  of  an  interior  daylight  distribution  with  arti- 

ficial light  sources    244 

Artificial    window    lighting 250 

A  cause  for  industrial  accidents 139 

Bedroom    259 

Causes  of  poor  illumination 275 

Characteristics   of   daylight   illumination 233,238 

Church    613 

A  classification  of  light  sources 236 

Color ;  its  use  in  lighting 663 

Colored  surroundings,  influence  on  the  color  of  the  useful  light    61 

Comparative  illuminations  produced  from  sources  of  the  same 

flux  by  direct,  indirect  and  semi-indirect  methods 118 

Concerning  data  on  competitive  illuminants 514 

Congruity   in   design    668 

Costs  for  lighting  of  various  interiors 672 

Cove  lighting   deficiencies    1 14 

Daylight :  the  ideal  illuminant 404 

Department  store  (see  heading:  Department  store  lighting  for 

sub-divisions)     '. 17 

Diffusion — growing  importance    661 

Dining-room    259 

Direction  of  light  for  interior  illumination.  .124,  126,  136,  260,  659 

Direction  and  distribution  of  daylight  in  interiors 113 

Direct  lighting  compared  with  indirect  and  semi-indirect 116 

Factories    470 

Gas  lighting.     (See  topics  of  gas  lighting.) 


XVI  TRANS.    I.  E.  S.      VOI,.  VIII 

Lighting  {continued):  page 

Of   home 149 

Home    experiments    in    illumination    from    large    area    light 

sources    229 

Hospitals    488 

Hygiene  and  safety  as  influenced  by  lighting 670 

Indirect  compared  with  direct  in  a  department  store 18 

Industrial    470 

Industrial   establishments    672 

Influence   of    decoration  and   contrast   in  the   illumination   of 

interiors    108 

Influence  of  good  lighting  upon  industrial  production 136,  287 

Importance  of   good  lighting 368 

Influence  of  light  between  6o°  and  horizontal  upon  sight 291 

Influence  of  surroundings   107 

Intensities    of    interior    illumination    compared    with    daylight 

intensities    112 

Interiors,  various  phases  and  problems 09 

Legislation    347 

Library    investigation    by    British    Illuminating    Engineering 

Society 344 

Location  of  light  sources  for  home  illumination 126 

Motion  picture  projectors    180 

Moving  picture   screens   and   auditoriums 187 

Policies  on  lighting  installations  of  different  companies 

481,    483,  5i6 

Its  principal  use  101 

A  psychological  aspect  284 

Progress 323,  345 

Railroad  passenger  cars   214 

Railway  cars    597 

Rug  racks   22 

Rural  districts,  sundry  systems 333 

Semi-indirect  from  side  walls 246,  261 

Simulation  of  daylight  for  interior  illumination 124 

Standards  compared  with  store  illumination 283 

Status  of  lighting  arc 652 

Steadiness  of  light,  its  importance 665 

Store.     (See  Store  lighting.) 

Street.     (See  topic  street  lighting.) 

-  A  Sunday-school  room    (gas) 439 

Tennis  courts    (by  gas) 347 

Tests  for  the  efficiency  of  the  eye  under  different  systems  of 

illumination 40, 

Typical    intensities,    store,    street,    show    windows,    factories, 

offices,  residences,  railway  cars 658 


SUBJECT   INDEX  xvii 

PAGE 
Lighting  Companies : 

Progressive  attitude  in  lighting  improvements  676 

Limen  test  for  color  sensitivity 41 

Luminosity : 

Theory  for  various  illuminants    407 

Lunch  room  lighting   508 

Magnetic  blow  in  arc  lamps  illustrated 165 

Magnetite  arc  lamps  increased  use 327 

Mantles :     (Gas.)     332 

Cause  of  falling  off  of  candle-power   332 

Developments   in   silk   mantles 332 

Theory  of  luminosity  of  Welsbach  mantle 407 

Mercantile  establishment  lighting   672 

Mazda  lamps.     (See  tungsten  lamps.) 

Mercury-vapor  lamps : 

"Reluctance"    explained    76,    86 

Theory,  operating  and  starting  characteristics  75,    86 

Voltage  required  depends  upon  vapor  pressure  76 

Disappearance  of  gas  due  to  chemical  action  rather  than  physi- 

cal   absorption    ; 330 

Developments   in   sealing    329 

Efficiency    of    mercury-vapor    lamp    combined    with    tungsten 

lamp,  light  to  give  white  light  328 

Characteristics  of  quartz  burner   84 

In  industrial  lighting  479,  487 

Millinery  store  lighting  519 

Moore   dioxid   tube   lamp   efficiency 376 

Motion  picture  lighting: 

Candle-power  performance  of   projectors   on   alternating  and 

direct    current     185 

Color  of  light  required 184,    195,  198 

Fatigue  and  eye  strain  185 

Intensity  of  illumination  on  screens  194 

Illumination  of  projectors    180 

Mechanism  of  lamps    182 

Projection  screens:  Advantages  and  disadvantages 186 

Nomenclature  and  standards.     (See  proposed  definitions,  page  6,  part 

1,  June  issue.) 

Nitrogen  filled  tungsten  lamp 325 

"Movies."     (See  motion  pictures.) 

Neon   tube   lamp    characteristics 330 

Short  life  due  to  absorption  of  neon  by  electrodes  330 

Characteristics  and  performance  of  lamps  371 


XV111  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 
New  York  City: 

Street   lighting   equipment    199 

Office  lighting : 

Average  illumination  intensity   658 

Parkway  lighting   90,  208 

Passenger  car  lighting:     (See  railing  car  lighting.) 

Pentane  lamp : 

Accuracy  values  in  photometry  419,  435 

Adjustment  of   lamps    415 

Candle-power   variations   and   peculiarities    434 

Details   of    operation    416 

Effect  of  pentane  density  on  candle-power  of  pentane  lamps..  425 

Effect  of  atmospheric  conditions  427 

General  directions  for  use 420 

Standardization  at  the  United  States  Bureau  of  Standards...  413 

Uses,  characteristics  and  deficiencies   41 1 

Variation  of  candle-power  with  height  of  flame 417 

Variation  of  candle-power  with  lumidity  and  barometric  pres- 

sure      429 

Variations  in  candle-power  with  different  grades  of  pentane  . .  436 

Photo-electric  cell : 

In   photometry    459 

Sensitivity    320,  468 

History    460 

Limitations   and   objections    465 

Methods  of  use  460 

Sensibility  curve 464 

Photography : 

In   illuminating   engineering    354 

Photometric  curves : 

Intensive   type   reflector    274 

Distributing  type  reflectors   274 

Extensive  type  reflector    271 

Focusing  type  reflector   274 

Photometry : 

Accuracy  in  studies    633,  648,  678 

Accuracy  of  Bunsen  and  Lummer-Brodhun  devices 634 

Average  error  of  test  plates 643 

Accuracy  variation  of  photometer  operators   647 

Accuracy  values  of  pentane  lamp   419 

Colorimeter,  use 71 

Colored  lights,  measurement.     (See  heterochromatic  below.) 

Description  of  new  photometers   339, 

Deficiencies  of  hemispherical  and  other  test  screens 158 

Error  due  to  parallelism  of  rays  of  reflected  light  sources 635 


Primary  standard  of  light,  progress  of  investigation  ....... . . .'  335 

-  Inadequacy  of  measurements  in  horizontal  plane  i58;  161 


SUBJECT   INDEX  xjx 


Photometry  (continued):  PAGE 

Errors   in   illumination   measurements  due   to    failure   of   test 
plate  to  obey  the  cosine  law 6 

Flicker  photometer    4" 

Hefner  as  a  standard   

Heterochromatic    ' 

Method   of   reducing  the   quantity   of   light  "for"  photometriS  3 
purposes    

Method  of  investigating  adjustment  errors  and' the 'calibration  3' 

of  portable  photometers   fi 

Method    of    measuring    the    energy    of    ultra-violet    "radiation 
emitted  by  mercury-vapor  lamp   

Pentane  lamp  as  a  working  standard  .,  4I0 

Planes  for  measuring  illumination  intensities  . . . .    '  228 

Photo-electric  cell,  use 
Primary  standard  of  lig 
Inadequacy  of  measurei 

Single  mirror  crane  type  . 

Selenium  cell,  use  and  characteristics   ' .'  33g 

Spectro-photometer,  use   

Street  illumination  tests  with  various  screens '.         I5- 

Comparison    of    tests    with    screens    of    various  "shapes  "and 

materials    , 

Use  of  helical  filament  lamps  as  standards 545',  548 

Pool  room  lighting  - 

Post  office  department,  railway  car  lighting  specifications .'.'  346 

Prismatic  glassware,  light  absorption  and  transmission 4S2 

Angle  of  prisms    

Distribution    \\  . 

Primary  standard  of  light;  progress  of  investigation 335 

Psychology : 

Of   illumination    fi 

Psychic  value  of  light,  shade,  form  and  color.........'.'..'.'.'.'.'  357 

Quartz  tube  lamp   R 

Efficiency    

*  "Z2Q 

Use  of  ultra-violet  rays  for  destruction  of  bacteria. . .         320 

For  street  lighting   "  32g 

Radiation : 

Of  carcel  lamp  , 

Efficiency  of  illuminants   '.'.'.'.'.    .' 40- 

Research  to  ascertain  effect  of  luminous  radiation  upon  the  eye  342 

—  Selective,  explained   .403,  & 


Railroad  signals  (visibility  of  signal  lights  of 


various  colors) 299 


XX  TRANS.    I.  E.  S.      VOI,.  VIII 

PAGE 
Railway  car  lighting: 

Car  lighting  specifications  of  the  post  office  department 346 

Center  deck  system    611 

Comparative  tests  with  various  luminants 217 

Cost  data  and  comparisons 599.  602 

Concentrated  filament  tungsten  lamps  for  locomotive  headlights  542 

Glare  from  tungsten  lamps 608,  609 

Headlights,   legislation    347 

History    215 

Illumination  test  data    597 

— —  Indirect   fixtures    348 

Intensities  of  illumination   223,  228 

Average  illumination  intensity 658 

Lighting  schedule    601 

Modern   practise    589 

Methods  of  producing  energy  for  illumination;  their  respective 

advantages  and  disadvantages    225 

Plans  for  photometering  illumination 228 

Progress  and  status   602 

Use  of  56-watt  and  94-watt  tungsten  lamps,  comparison 

592,    604,  606 

Wiring    227,  593 

Tests  in  New  York  subway 606 

Reflection : 

Classes  of  reflecting  surfaces 270 

Coefficients    69 

Diffuse    220 

Influence  of  colored  surfaces  on  light  reflected 62 

Influence   of   character   of   reflecting   surfaces   on   the   design 

and  light  distribution  of  reflectors 273 

Law  of,  explained  and  illustrated 270 

Regular    270 

Specular    270 

Specular  influence  on  light  reflected 67 

Reflectors:     (See  also  prismatic.) 

Aluminum    282 

Aluminum  and  porcelain  enameled  finishes  for  reflectors  com- 

pared   281 

Advantages  of  reflectors  of  various  metals 269 

A  reflector   for  copying  the  distribution  of   daylight   from   a 

window    248 

Absorption  of  light 107,  281,  290,  452, 

Absorption  of  light  by  aluminum  finished  reflectors 290 

Basis   for  comparing  efficiencies 28a 

Characteristics  of  various  metal  reflectors 271 


SUBJECT   INDEX  xxi 

Reflectors  {continued):  page 

Considerations  which  govern  selection  for  an  installation 277 

Cost  considerations  in  the  design  of  lighting  installations 282 

Efficiency  of   several   types 289 

Elimination  of  streaky  illumination  effect  from  metal  reflectors  286 

Glass  vs.  metal  for  industrial  lighting 288 

Historical  notes  on  the  design  of  metal  reflectors 285 

Influence    of    shape    of    reflectors    on    light    distribution    and 

efficiency    289 

Influence  of  deep  reflectors  on  life  of  electric  lamps 288 

Influence  of  character  of  reflecting  surfaces  on  the  design  and 

light  distribution  of  reflectors 273 

Light  distribution  depending  upon  design 105 

Photometric  curve  of  extensive  type  reflector 274 

Photometric  curve  of  distributing  type  reflectors 274 

Photometric  curve  of  intensive  type  of  reflectors 274 

Porcelain   enamel,   characteristics    282 

Proposal  to  use  1,000  lumen  basis  in  making  distribution  curves 

from   reflectors    645,  647 

Spacing  and  mounting  heights  for  various  types  of  reflectors..  272 

Uses  and  purposes   104 

Regular  reflection : 

Explained   and   illustrated    270 

Residence  lighting: 

Average  illumination  intensity   658 

Cost  in  small  homes 672 

Restaurant  lighting    527 

Rug  rack  lighting   22 

Selective   radiation   explained    403,  406 

Seeing:     (See  also  visual  acuity.) 

Influence  of  glare  104 

Efficiency  under  lights  of  various  colors 144 

Selling   illumination    499 

Semi-indirect  lighting : 

Comparative  illuminations  produced  from  sources  of  the  same 

flux  by  direct,  indirect  and  semi-indirect  methods 118 

Compared  with  direct  and  indirect 116 

By  gas  in  an  exhibition  hall 263 

In    churches    616 

From    walls    .' 246,  261 

Shoe  store  lighting    508,  522 

Show  case  lighting  30 

Show  window  lighting.     (See  window  lighting.) 

Signals : 

Visibility  of  railroad  signals  of  various  colors 299 

Perception  of  lights  of  short  duration 342 

Use  of  concentrated  filament  lamps 542 


xxii  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 

Sign  lamps    (tungsten    324,  539 

Slagging  of  flame  carbon  electrodes ;  cause  explained 171 

Store    (small)   lighting ^72 

Snellen  type  test  for  efficiency  of  the  eye 42 

Spacings : 

And  heights  for  various  types  of  reflectors 272 

Spectrum : 

Of  acetylene   4°5 

Use  in  determination  of  value  of  units 4°7 

Specular   reflection    667 

Influence  on  color  of  light  reflected  from  colored  surfaces 67 

— ■—  Explained  and  illustrated   270 

Standards : 
The  pentane  lamp  in  photometry   4xo 

The  Hefner  lamp  412,  422 

Primary,  progress  of   investigation  of    335 

Store  lighting: 

Antique  and  curio  shop  524 

Carpet  store    5<>8 

Department:     (McCreery  &  Co.,  Pittsburgh.)  See  also  depart- 

ment store  lighting  17 

Barber    shop    508 

Books    524 

Candy 520 

Delicacies    528 

Grocery  store    521 

Jewelry     508,  517 

Ladies    wear    525 

Lunch    room    508 

Millinery     .  519 

■  Pool  room    5°8 

Restaurant    527 

Shoe    508,  522 

Stationery    526 

Standards  compared  with  store  illumination    283 

— —  Tailor-shop     508 

Tea    room    522 

■  Toggery  or  haberdashery  shop 5J8 

Toy    517 

Street  car  lighting.     (See  railway  car  lighting.) 

Street  lighting: 

Average  and  typical  illumination  intensity  658 

■  Business   streets    requirements    88 

Color  of  light  required  89 


SUBJECT   INDEX  xxiii 

PAGE 
Street  lighting  {continued): 

■  Controversy    over   an    installation    of    high    pressure   gas    and 

flaming  arc  lamps  in  England  349 

Discernment  of  objects  by  their   silhouettes 94 

Flame  arc  lamps  in  Chicago 485 

Of  high  pressure  gas  lighting,  progress  and  status 350 

Intensity  required    89 

Inadequacy  of  photometrical  measurements  in  horizontal  plane  158 

New  York  City   199 

Ornamental  lamp-posts    92 

Height   of   flame   arc   lamps 485 

Photometry  with  various  screens    155 

Progress  and  status   348 

Requirements    88 

Spacing  and  height  of  illuminants 92,  95,    96 

Residence  streets  requirements    90,    97 

Status  in  France   351 

With  ornamental  arc  lamps  88 

Sunday  school  room  lighting  with  gas   439 

Sunlight,  intensity  compared  with  artificial  illumination   301 

Tantalum   lamps,   relative  number   of   carbon  gem,   tantalum   electric 

incandescent  lamps  sold  in  1907,  1908,  1909,  1910,  191 1,  1912.  323 

Tailor   shop   lighting    • 508 

Tea  room  lighting 522 

Tennis  court  lighting   (gas) 347 

Tests : 

For  brightness   discrimination    43 

For  the  efficiency  of  the  eye  under  different  systems  of  illu- 

mination      40 

Limen  test  for  color  sensitivity 41 

Snellen  type  test  for  the  efficiency  of  the  eye  43 

Visual    acuity    43 

Railroad  car  lighting   217 

Theatre  lighting : 

Color  of  light  for  make-ups   35 

Titanium    arc   lamps,    characteristics 407 

Toy   Store    517 

Train  lighting.     (See  railway  car  lighting.) 

Trolley  car  lighting.     (See  railway  car  lighting.) 
Tungsten  lamps : 

Blackening  of  bulbs,  cause  of  549 

Bulb  size,  decrease   536 

Candle-power  change  with  variation  in  voltage 595 

Brightness  of  filament  233,  353,  483 


XXIV  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 
Tungsten  lamps  {continued) : 

Characteristics  of  tungsten  filament  railway  lamps  592 

Cooling  effect  the  leading-in  wires  upon  filaments  of  lamps  of 

street  series  types   385 

Concentrated    filament    325,  537.  54*.  543 

Discoloration  of  lamp  bulbs  due  to  chemicals,  and  its  effect 

upon   candle-power    546 

Daylight  approximation  lamp   326 

Effect  of  frosting  on  effective  light  288 

Effect  upon  life,  of  enclosing  glassware   546 

Efficiencies     324 

— —  Focusing  lamp,  construction  and  efficiency  326 

Helical  filaments    537 

Helical  filament  lamps  as  standards  in  photometry 545,  548 

Improvements  in  candle-power  performance  534 

Influence  of  deep  reflectors  on  life 288 

Intrinsic  brilliancy  of  filament 233,  353,  483 

Life  and  candle-power  tests 543,  550,  553 

Light  of  tungsten  lamps  vs.  carbon  lamps  for  seeing 146 

Miniature    construction    details    325 

Method  of  testing  for  spots  in  incandescent  lamps 326 

Patented  bulb  for  reducing  glare  of  filament 327 

Percentage  of  incandescent  lamp  sales  for  3  months  of  19 13..  656 

56  and  94-watt  railway  tungsten  lamp  systems  compared 

594,    604,  606 

Recent  improvements    324,  553 

Use  of  chemicals  to  prevent  bulb  blackening 544,  547,  549 

Relative  number  of  carbon,  gem,  tantalum,  electric  incandes- 

cent lamps  sold  in  1907,   1908,  1909,  1910,  191 1,   1912 223 

Reducing  blackening  of  bulbs  with  "vacuum  getter" 324 

Sign   lamps 324 

Smashing  point    540,  546,  547 

Standardizations    536 

Variation  in  candle-power  produced  by  bulb  frosting 326 

vs.  arc  lamps  in  a  department  store 17 

Ultra-violet  light : 

"Chinese  white"  test  for  presence 355 

Effect  upon  the  eye  of  ultra-violet  light  emitted  by  commer- 

cial light  sources  343 

From   artificial   light   sources 132 

Influence   on  the   eye 133,  138 

Media  opaque  and  transparent  to  ultra-violet  radiation 140 


SUBJECT    INDEX  XXV 

PAGE 
Ultra-violet  light  {continued): 

- A   method  of   measuring  the   energy  of   ultra-violet   radiation 

emitted  by  mercury-vapor  lamp 340 

Radiated   from  arc  lamps 165 

Use  of  rays  from  quartz  tube  lamp  for  destruction  of  bacteria  329 

Vacuum  tube  lamps : 

Cadmium  vapor  lamp    329 

Developments  in  sealing  329 

Efficiency  of  Neon  and  Moore  tube  lamps 376 

Disappearance    of    gas    due    to    chemical    action    rather    than 

physical    absorption    330 

Mercury-vapor    329 

Neon  tube  lamp   330,  376 

Vision.     (See  visual  acuity.) 
Visual  acuity: 

And  monochromatic  light    298 

Influenced  by  brightness  of  surroundings J92,  297,  298,  299 

Influence  of  light  between  60  °  and  horizontal  upon  sight 291 

Influence  of  angle  of  direction  of  light  upon  efficiency  of  eye..  277 

Tests  for  the  eye 43 

Tests  of   legibility   of  type 342 

Under  light  of  various  colors 145 

Vision  influenced  from  color  of  light 143 

Visibility  of  signal  lights  of  various  colors 299 

Window  lighting: 

Average  illumination  intensity   658 

Department  store   28,  557 

Intensity  of   illumination   required 89 

Selection  and  spacing  of  reflectors 573 

Specifications,  requirements  and  tests 557 

Walls : 

Brightness  an  element  affecting  ocular  comfort 108 

Welsbach  mantle : 

■  Intrinsic   brilliancy    233 

Zeiss    refractometer    for    determination    for    CO-    content    in   atmos- 
phere   414,  435 

Zeiss  system  of  illumination  intensities  for  hospitals 497 


INDEX  TO  AUTHORS 


The  letter  d  indicates  discussion. 

PAGE 

Alger,  Ellice  M.  (M.  D.).     Illumination  and  eyestrain 130 

d — Eyestrain  and  motion  pictures 190 

Amrine,  T.  H.     The  cooling  effect  of  leading-in  wires  upon  the  fila- 
ments of  tungsten  incandescent  lamps  of  the  street  series 

type    385 

d — Accuracy  in  photometry  with  Bunsen  and  Lummer-Brodhun 

photometer    647 

Baldwin,  A.  T.,  W.  R.  Mott  and  R.  B.  Chillas,  Jr.    d — Flame  arc 

lamp    carbons    177 

Barrows,    George    S.     d — Co-operation   of   architect,    decorator   and 

illuminating   engineer   in   illumination   problems 153 

Bettcher,  C.  W.     d — Candle-power  variation  of  tungsten  lamps  due 

to  change  in  voltage 609 

Bond,  C.  O.     d — Test  for  the  efficiency  of  the  eye 59 

d — Home  lighting    261 

d — Use  of  reflectors  in  railroad   lighting 284 

CalvERT,    H.     d — Industrial   lighting 283 

d — Hospital   ward    lighting    497 

d — Central  station  free  renewal  policy 543 

CauldwEll,  F.   C.     d — Church  lighting 625 

Chapin,   H.   C.     d — White  and  yellow   flame  carbon   arc  lamps    for 

factory   lighting    484 

Chillas,  R.  B.,  Jr.,  W.  R.  Mott  and  A.  T.  Baldwin,    d — Flame  arc 

lamp  carbons    177 

Claude,  George.     Neon  tube  lighting 371 

ClEwell,  Clarence  E.     d — Street  lighting 212 

Cobb,    Percy   W.     Vision   as    influenced    by   the    brightness    of    sur- 
roundings     292,  299 

d — Effect  of  illumination  on  the  eye 138 

d — Illumination  intensity  and  motion  pictures 194 

Cotton,  A.  C.     d — Illumination  of  passenger  cars 225 

CowlES,  J.  W.     d — Street  lighting 212 

Cravath,  J.  R.     d — Influence  of  surroundings  on  vision 297 

d — Show  window  lighting    s 587 

d — Church   lighting    623,  627 

Crittenden,  E.  C,  and  A.  H.  Taylor.    The  pentane  lamp  as  a  work- 
ing standard    410 

Darrah,  W.  A.    The  flame  carbon  arc  lamp 162 

Some  theoretical  considerations  of  light  production _400 

Dicker,  A.  O.,  and  M.  H.  FlExner.    Factory  lighting 470,  485 


INDEX    TO   AUTHORS  xxvii 

PAGE 
Dunning,  H.  S.     d — Effect  of  inclosing  globes  on  life  of  tungsten 

lamps    456 

d — Candle-power  performances  of  tungsten  lamp 546 

Edwards,  Evan  J.,  and  Ward  Harrison.     Recent  improvements  in 

incandescent  lamp  manufacture   533 

Some  studies  in  accuracy  of  photometry 633.  649 

Edwards,  Evan  J.     d — Temperature  characteristics  of  electric  incan- 
descent lamp  filaments    308 

Ely,   Robert  B.     Church  lighting 613,  630 

d — Home  illumination    259 

d — Factory   lighting    483 

d — Store  lighting  531 

d — Illumination  equipment  of  street  cars 610 

Fabry,  C.  H.    A  practical  solution  of  the  problem  of  heterochromatic 

photometry    302 

FerreE,  C.  E.    Test  for  the  efficiency  of  the  eye  under  different  sys- 
tems   of    illumination    and    a    preliminary    study    of    the 

causes  of  discomfort  40,    60 

d — Illumination  and  visual  acuity 142 

d — Illumination  in  the  home  from  large  area  light  sources 255 

Flexner,  M.  H.,  and  A.  O.  Dicker.     Factory  lighting 470,  485 

d — Chicago  street  lighting  with  flame  arc  lamps 485 

Ford,  Arthur  H.     A  photometer  screen  for  use  in  tests  of  street 

illumination    155 

Gage,  H.  P.    d — Surroundings  and  vision ;  monochromatic  light ;  rail- 
road signals    298 

Graves,    C.   B.     d — Indirect   lighting 128 

Halvorson,  C.  A.  B.,  Jr.     Street  lighting  with  ornamental  luminous 

arc  lamps    88 

Harrison,   Ward,  and  Evan  J.  Edwards.     Recent  improvements   in 

incandescent  lamp  manufacture    533,  547 

Some   studies   in   accuracy   of   photometry 633 

Harrison,  Ward,     d — Department  store  lighting 37 

d — Competitive  tests  of  store  lighting 514 

d — Wiring  of  street  railway  cars 610 

Hibben,  S.  G.     Modern  practise  in  street  railway  illumination. ..  .589,  610 

d — Department  lighting    36 

d — Influence  of  enclosing  glassware  on  light  distribution 496 

d — Store  lighting    529 

HoadlEy,  Geo.  A.     Indirect  illumination  on  the  eye 121 

d — Home  lighting    254 

d — Industrial   lighting   reflectors    284 

Howell,   J.   W.     d — Recent   improvements    in   tungsten   incandescent 

lamps    549 

Hunter,  George  Leland.    Home  lighting 149 


XXVlii  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 
Israel,  Joseph  D.     Annual  report  of  the  General  Secretary  for  the 

fiscal  year  ending  September  30,    1913 683 

Ives.  Herbert  E.    Some  home  experiments  in  illumination  from  large 

area  light  sources    229,  258 

d — Tests  of  visual  efficiency 57 

d — Lighting  interiors    122 

d — Influence  of  surroundings  on  vision 298 

d — Heterochromatic  photometry    319 

d — Interior   illumination    - .  369 

d — Use  of  photo-electric  cell  in  photometry 467 

d — Uses  of  the  helical  filament  lamp 544 

Jackson,  J.  B.    d — Cost  of  lighting  units  for  street  car  service 610 

Josselyn,  A.  E.    d — Some  causes  of  poor  illumination 120 

Kiefer,  L.  T.     d — Showcase  lighting 37 

Kilmer,  William   S.     Hospital  lighting 488 

Kingsbury,  Edwin  F.    Experiments  in  the  illumination  of  a  Sunday- 
school  room  with  gas 439 

Lacombe,  C.  F.     Street  lighting  of  greater  New  York 199 

Lansingh,  V.  R.     Characteristics  of  enclosing  glassware 447 

d — Uses  of  the  helical  filament  lamp t . . .  545 

d — Railway  car  illumination   609 

Law,  Clarence  L.,  and  A.  L.  Powell.    Distinctive  store  lighting 515 

LEE,  J.  W.     d — Glass  and  metal  reflectors  for  industrial  lighting....  285 
Lewinson,  L.  J.     d — Improvement  in  the  candle-power  performance 

of  tungsten  lamps  from  191 1  to  1913 552 

d — Variations   in  accuracy  of   photometry   with   different  pho- 
tometers      647 

Lewis,  F.  Park    (M.  D.).     The  physic  values  of   light,  shade,  form 

and  color    357 

LitlE,  T.  J.,  Jr.     d — Pressure  of  gas  lighting  for  stores 513 

d — Church  lighting    628 

Little,  W.  F.     d — Metal  reflectors  for  industrial  lighting 290 

d — Hospital  lighting   497 

d — Standardization  of  incandescent  lamp  filament  dimensions..  541 
Lloyd,   M.  G.     d— Temperature  characteristics   of  a  helical  filament 

lamp   543 

d — Smashing  point  of  tungsten  lamps 546 

Lloyd,  R.  L.     d— Direction  of  light  for  interior  illumination. 260 

Lloyd,  E.  W.     d — Factory  lighting 484 

Luckiesh,  M.     The  influence  of- colored  surroundings  on  the  color 

of    the    useful    light 62 

d — Measurement   of   illumination   efficiency 159 

d — Influence  of  surroundings  on  vision 299 

d — Temperature    and    candle-power    characteristics    of    electric     - 
incandescent  lamp  filaments 399 


INDEX    TO    AUTHORS  XXIX 

PAGE 
Luckiesh,  M.  {continued) 

d — Church  lighting   629 

d — Variations    in    accuracy    of   photometry 648 

McAllister,  A.  S.     d — Direction  of  light  for  interior  illumination..  126 
Macbeth,  Norman,     d — Lack  of  agreement  in  standard  specifications 

of  incandescent  lamps   540 

Madgsick,   H.   H.     d — Factory  lighting 485 

Marks,    L.    B.     d— Daylight    illumination 124 

d — Influence  of  surroundings  on  vision 297 

Martin,  J.  Frank.     The  illumination  of  motion  picture  projectors..  180 
Millar,    Preston    S.      Progress    and    functions    of    the    Illuminating 

Engineering   Society    (inaugural  address) 1 

Report  of  the  general  secretary  for  1912 6 

Some  phases  of  the  illumination  of  interiors 99 

The  status  of  the  lighting  art    (presidential  address) 652 

d — Illumination  of  passenger  cars 224 

Minick,  J.  L.     Illumination  of  passenger  cars 214,  226 

Moore,  D.  McFarlan.    d — Large  area  light  sources 126 

Mott,  W.  R.,  R.  B.  Chillas,  Jr.,  and  A.  T.  Baldwin,  d — Flame  arc 

lamp   carbons    177 

Myers,  R.  E.    d — Improvements  in  the  manufacture  of  tungsten  lamps  544 
Nichols,  G.  B.     d — Prevention  of  poor  illumination  by  co-operation 

of  illuminating  engineer  and  architect 153 

Philbrick,  J.  E.     d — Store  lighting 499 

Pierce,  Robert  F.     Gas  lighting  in  an  exhibition  hall 263 

d — Store  lighting  511 

d — Church  lighting    625 

Porter,  L.  C.     d — Passenger  car  lighting 227 

d — Uses  of  the  concentrated  tungsten  filament  lamp 542 

d — Tungsten   lamps    for   railway   cars 604 

Powell,  A.  L.,  and  Clarence  L.  Law.    Distinctive  store  lighting.  .515,  531 

Powell,  A.  L.     d— Church  lighting 623 

Reid,  H.  A.     d — Cost  and  depreciation  of  tungsten  lamps  and  reflec- 
tors  for  factory  lighting 482 

Richtmyer,  F.  K.     The  photo-electric  cell  in  photometry 459 

Rolph,  Thomas  W.    Metal  reflectors  for  industrial  lighting.  .268.  287.  289 

Roosa,  G.  W.     d — Flame  carbon  lamps  for  industrial  lighting 483 

Rose,  S.  L.  E.     The  illuminating  engineering  laboratory  of  the  Gen- 
eral Electric  Company  379 

d — Efficiency  of  six  ampere  flame  arc  lamps 484 

d — Plotting    photometric    distribution    curves    of    incandescent 

lamps    647 

Rows.  E.  B.     d — Interpreting  illumination  test  data 514 

d — Tungsten  lamp  wiring  and  illumination  for  railway  cars....   605 
d — Church  lighting    628 


XXX  TRANS.    I.  E.  S.      VOL.  VIII 

PAGE 

Rowland,   A.  J.     d — Home  lighting 260 

d — Development  of  the  metal  shade  for  industrial  lighting....  284 

Shalling,  H.  W.     Department  store  lighting 17,  80 

d — Department  store  lighting   38 

Sharp,    C.    H.      d — Daylight    illumination 123 

d — Photometry  of  street  illumination 158 

Simon,  Edward  L.     d — Illumination  and  motion  pictures 197 

Skiff,   W.    M.     d — Showcase   lighting 35 

Smith,  Louis  C.     d — Illumination  and  motion  pictures 197 

Stark,    A.    W.      d — Interior    illumination 128 

SticknEy,  G.  H.     d — Department  store  lighting 33 

d— Measurement  of  illumination   161 

d — Concentrated  filament  lamp    541 

d — Railway  car  illumination   608 

d — Church  lighting   629 

Taylor,  A.  H.,  and  E.  C.  Crittenden.    The  pentane  lamp  as  a  work- 
ing standard   410 

Thomas,  Percy  H.    Theory  of  mercury-vapor  apparatus 75 

Tousey,  Sinclair,     d — Illumination  and  eyestrain 140 

Vaughn,  F.  A.     d — Eyestrain  and  motion  pictures 192 

Ware,  R.  C.     d — Poor  illumination  and  bare  lamps 120 

Wheeler,  H.  B.     d — Hospital  lighting 498 

The  lighting  of  show  windows 555,  588 


m} 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 

Published  on  the  2Sth  of  each  month,  except  during  Ju!y,  August,  and  September,  by  the 

ILLUMINATING   ENGINEERING  SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIII 


JANUARY.  1913 


No.  1 


Index  for  Volume  VII. 

The  index  for  Volume  VII  (1912) 
of  the  Transactions  will  be  mailed  in 
separate  form  with  the  February  issue 
which  will  be  out  about  the  middle  of 
March. 


The  New  Transactions. 

In  this  issue  an  attempt  has  been 
made  to  improve  the  general  make-up 
of  the  Transactions.  Dull  finished 
paper,  free  from  the  objectionable  glare 
which  a  reader  usually  encounters  in 
magazines  and  books,  has  been  used 
throughout.  This  number  also  has  a 
more  appropriate  cover.  Other  im- 
provements, in  the  way  of  better 
arrrangement  and  presentation  of  sub- 
ject matter,  more  legible  type,  and  the 
like,  may  be  expected  later.  Such 
changes  will  ultimately  afford  a  more 
commendable  publication  which,  it  is 
hoped,  will  be  of  greater  interest  to  the 
members  of  the  society  and  more  val- 
uable as  a  journal  of  reference  for 
libraries.  Beginning  with  the  April 
issue  each  number  of  the  Transactions 
will  be  published  on  the  28th  of  the 
month. 


Annual  Committee  Reports. 

Below  are  given  synopses  of  the 
annual  reports  of  committees,  which 
were    presented    at    a    meeting    of    the 


council  held  in  New  York,  January  10, 
I9I3- 

finance  committee 
All  the  bills  for  the  year  were  ap- 
proved by  the  committee.  A  report 
on  the  books  of  account  for  1912  pre- 
pared by  Wm.  J.  Struss  &  Company, 
certified  public  accountants,  accom- 
panied the  committee's  report.  This  re- 
port showed  a  deficit  of  $640.07  for  the 
year,  and  an  impairment  of  surplus 
amounting  to  $1,435-63-  The  average 
expense  per  member  was  estimated  as 
$7.36.  The  report  appears  in  full  else- 
where in  this  issue  of  the  Transactions. 

COMMITTEE    ON    ILLUMINATION 
PRIMER 

This  report  is  supplementary  to  the 
report  of  the  committee  in  the  June 
(1912)  Transactions.  Ten  thousand 
primers  (Light:  Its  Use  and  Misuse) 
had  been  published,  and  an  edition  of 
5,000  is  now  on  the  press.  About  5,500 
copies  have  been  distributed  free  by  the 
society,  4,000  have  been  sold  and  deliv- 
ered, and  orders  are  on  hand  for  4,500. 
Forty-five  lighting  companies  scattered 
throughout  the  country  have  purchased 
copies  mostly  in  lots  of  25,  50  and  100. 
Orders  have  been  received  from  a 
number  of  manufacturers  of  lamps  and 
lighting  appliances,  from  contractors  and 
from  several  colleges.  The  primer  was 
reprinted  in  full  in  journals  whose  com- 
bined circulation  is  about  30,000;  ab- 
stracts have  appeared  in  more  than  150 


TRANSACTIONS   I.    E.    S. — PART   I 


domestic  and  foreign  journals;  it  was 
estimated  that  altogether  about  2,000,000 
notices  of  the  primer  have  been  printed 
by  these  journals.  The  primer  was 
published  in  England  in  a  modified 
form.  A  much  wider  and  more  general 
distribution  or  circularization  is  very 
probable. 

COMMITTEE    ON    FACTORY    LIGHTING 
LEGISLATION 

The  committee,  although  it  had  been 
appointed  recently,  drafted  and  sub- 
mitted to  the  New  York  Investigating 
Commission  recommendations  on  sec- 
tions 3  and  4  of  proposed  Bill  No.  18 
pertaining  to  the  lighting  of  factories 
and  work-rooms.  These  recommenda- 
tions having  to  do  with  specifications 
for  proper  and  adequate  lighting  were 
made  principally  in  the  interests  of 
ocular  hygiene  and  safety  of  employees. 
The  bill  in  its  final  form  will  be  printed 
in  a  future  issue  of  the  Transactions. 

COMMITTEE  ON    GLARE  FROM   REELECT- 
ING SURFACES 

This  committee  had  been  in  existence 
only  a  short  time.  In  the  course  of  its 
work  it  learned  (1)  that  school-book 
publishers  are  in  favor  of  eliminating 
glazed  papers;  but  they  require  for 
books  a  cheap  and  durable  paper 
which  will  reproduce  half-tones  well ; 
(2)  paper  manufacturers  contend  that 
they  will  produce  a  paper  to  meet  the 
requirements  of  publishers  when  there 
is  a  demand  for  it;  (3)  school  officials 
have  given  little  or  no  attention  to 
glazed  surfaces.  The  committee  has 
contemplated  conducting,  in  conjunction 
with  the  research  committee,  an  inves- 
tigation of  the  question  of  glare  from 
paper;  it  proposed  the  starting,  at  some 
future  date,  of  a  definite  movement  to 
eliminate  polished  surfaces  wherever 
possible,  particularly  glazed  paper  from 
school  books. 


COMMITTEE  ON   RECIPROCAL  RELATIONS 
WITH   OTHER  SOCIETIES 

Co-operative  relations  with  some 
thirty-two  professional,  scientific,  com- 
mercial and  philanthropic  organizations 
were  promoted  during  the  year.  Joint 
meetings  were  held  with  a  number  of 
others.  "It  is  believed  that  through  the 
work  of  the  committee  the  society's 
name  and  influence  have  been  materially 
advanced.  The  I.  E.  S.  has  been 
brought  to  the  attention  of  various 
organizations,  heretofore  ignorant  or 
indifferent  to  its  existence.  While  in 
every  case  we  have  not  met  with  the 
success  we  had  hoped,  we  believe  our 
work  to  be  cumulative  and,  if  properly 
followed  up,  will  ultimately  be  produc- 
tive of  most  profitable  relations  with 
all  organizations  interested  in  illumina- 
tion." The  report  also  contained  sug- 
gestions for  the  next  committee.  The 
committee  promoted  co-operative  rela- 
tions with  the  following  societies : 

American  Academy  of  Medicine. 

American  Association  of  Cotton  Man- 
ufacturers   (Boston,  Mass.). 

American  Association  of  Cotton  Man- 
ufacturers  (Charlotte,  N.  C). 

American  Electro-Therapeutic  Asso- 
ciation. 

American  Gas  Institute. 

American  Institute  of  Architects. 

American  Institute  of  Mining  Engi- 
neers. 

American   Ophthalmological    Society. 

American   Medical  Association. 

American   Public   Health  Association. 

The  American  School  Hygiene  Asso- 
ciation. 

American  Society  of  Mechanical 
Engineers. 

American  Association  for  Conserva- 
tion of  Vision. 

Architects  and  Engineers  Club.- 

Architectural  Club  of  Washington. 


TRANSACTIONS   I.    E.    S. — PART   I 


Association    of    Edison    Illuminating 

Companies. 
Association   of   Iron   and   Steel   Elec- 
trical Engineers. 
Association     of     Railway     Electrical 

Engineers. 
Association  of   Railway   Surgeons. 
Association   of    Stationary   Engineers. 
Committee     on     the     Prevention     of 

blindness. 
Industrial   Safety  Association. 
Institute  of  Electrical  Engineers. 
Medical     Society     of     the     State     of 

Illinois. 
Medical  Society  of  the  State  of  New 

York. 
Medical     Society     of     the     State     of 

Pennsylvania. 
Museum  of   Safety. 
National  Commercial  Gas  Association. 
National   Electric  Light  Association. 
Ohio  State  Medical  Association. 
Physiological   Society. 

COMMITTEE   ON    PROGRESS 

This  report  was  supplementary  to  the 
report  of  the  committee  which  is  printed 
in  the  November  issue  of  the  Trans- 
actions ;  it  suggests  that  it  would  be 
desirable  for  the  next  progress  com- 
mittee to  endeavor  to  base  much  of  its 
report  on  a  review  of  articles  pertain- 
ing to  illuminating  engineering  which 
may  appear  in  the  various  domestic  and 
foreign  journals;  in  other  words,  a 
report  should  constitute  as  far  as  pos- 
sible a  conspectus  of  progress  in  illumi- 
nating engineering.  The  preparation  of 
such  a  report,  it  was  stated,  would  in- 
volve a  vast  amount  of  painstaking 
work  in  abstracting  and  indexing  matter 
from  many  publications. 

SECTION     DEVELOPMENT    COMMITTEE 

During  the  year  the  committee  held 
two  meetings ;  recommended  the  ap- 
pointment of  section  representatives  on 
the   national   papers   committee;    under- 


took the  preparation  of  a  guide  on  sec- 
tion management;  and  suggested  the 
appointment  of  representatives  or  local 
secretaries  in  cities  not  having  sections ; 
and  in  a  general  way  sought  to  promote 
the  welfare  of  the  sections  and  the 
society.  The  appointment  of  a  commit- 
tee to  organize  a  section  comprising 
cities  in  the  Lake  Erie  region  was  made 
as  a  result  of  a  recommendation  of  the 
committee. 

COMMITTEE    ON     NOMENCLATURE    AND 
STANDARDS 

The  report  appears  in  full  in  the 
December  issue  (1912)  of  the  Trans- 
actions; it  gives  a  number  of  new 
photometric  definitions  and  a  brief  ac- 
count of  the  committee's  activities. 

COMMITTEE  ON  RESEARCH 
The  report  stated  that  the  committee 
hoped  to  be  able  to  submit  a  report  of 
accomplishments  next  year,  .if  the  coun- 
cil deems  it  advisable  to  continue  the 
committee.  The  functions  of  the  com- 
mittee were  outlined  as  follows :  "It 
should  be  a  sort  of  a  clearing-house  in 
illuminating  engineering  matters.  It 
should  endeavor  to  see  that  investiga- 
tions which  are  necessary  for  applica- 
tion in  practise  are  undertaken  by  those 
competent  and  prepared  to  do  such 
work.  It  should  be  prepared  to  recom- 
mend to  those  desiring  to  undertake 
research  work  suitable  problems  to  be 
investigated.  It  should  bring  into  closer 
co-operation  the  various  scientific  and 
technical  bodies  and  should  keep  in  close 
touch  with  the  various  scientific  and 
technical  schools  where  research  work 
in  any  of  the  allied  sciences  is  under- 
taken." 

COMMITTEE   ON   EDITING    AND 
PUBLICATION 

The  Transactions  for  1912  had  ap- 
proximately   300    pages    less    than    the 


TRANSACTIONS   I.    E.    S. — PART   I 


Transactions  of  191 1,  and  the  cost  was 
about  $1,100  less.  Condensation  and 
elimination  accounted  for  practically  all 
of  both  these  reductions.  Three  recom- 
mendations were  made:  (1)  the  ar- 
rangement of  the  Transactions  into 
two  parts — a  news  section  and  a  section 
devoted  entirely  to  papers,  discussion 
and  reports;  (2)  the  use  of  a  paper 
freer  from  glare  than  the  paper  in  use; 
(3)  the  publication  of  a  guide  setting 
forth  the  requirements  and  general 
style  of  papers  and  discussion  which 
would  reduce  printing  expenses  and 
facilitate  the  work  of  publication. 

COMMITTEE    ON    ADVERTISING 

The  advertising  revenue  of  1912  was 
$1,356,  as  compared  with  $1,225  for  1911. 
Contracts  obtained  for  additional  space 
in  the  Transactions  together  with 
those  pending  should  net  about  $2,500 
in   1913. 

COMMITTEE     ON     NEW     MEMBERSHIP 

A  quiet  but  conservative  campaign 
was  made  for  new  members.  Most  of 
the  201  applications  and  3  reinstate- 
ments during  the  year  1912  are  attrib- 
utable to  the  activities  of  the  commit- 
tees. The  committee  recommended  that 
the  next  committee  be  made  up  of  the 
chairmen  of  the  section  membership 
committees  and  such  others  as  may  be 
expedient  in  the  conduct  of  the  com- 
mittee's work.  If  the  committee  will 
keep  in  close  touch  with  the  section 
membership  committees  and  encourage 
an  exchange  of  ideas  or  plans,  it  was 
said  that  its  work  will  be  greatly  facili- 
tated. 

1912  CONVENTION   COMMITTEE 

This  report  was  the  last  of  a  series 
of  reports  by  the  committee.  It  in- 
cluded a  scrap  book  of  samples  of  all 
the  literature  of  the  recent  convention 
and   a  number  of   valuable   suggestions 


as  to  procedure,  which  should  be  of  use 
to  future  convention  committees.  A 
check  for  $177.74,  the  excess  of  receipts 
over  expenditures  by  the  committee, 
accompanied  the  report. 


Council  Notes. 

JANUARY    COUNCIL    MEETING 

Twenty-five  applicants  were  elected 
to  membership  at  a  regular  meeting  of 
the  council  which  was  held  in  the  gen- 
eral offices  of  the  society,  29  West 
Thirty-ninth  Street,  New  York,  Jan- 
uary 10,  1913.  The  names  of  those 
new  members  are  listed  on  page  5. 
Twenty-nine  resignations,  most  of 
which  had  been  held  over  from  the  pre- 
vious year,  were  accepted. 

A  series  of  amendments  to  the  by- 
laws, most  of  which  were  necessitated 
by  a  recent  constitutional  change  in  the 
fiscal  year  of  the  society,  was  read  for 
the  first  time. 

Annual  reports  were  received  from 
the  following  committees :  new  mem- 
bership, finance,  nomenclature  and 
■  standards,  editing  and  publication,  re- 
search, reciprocal  relations  with  other 
societies,  section  development,  glare 
from  reflecting  surfaces;  progress, 
illumination  primer,  factory  lighting 
legislation.  A  synopsis  of  each  report 
is  given  elsewhere  in  this  issue. 

The  final  report  of  the  1912  conven- 
tion committee  was  received. 

President  Lansingh  reported  progress 
in  the  work  of  organization  of  a  Lake 
Erie  section  of  the  society. 

Mr.  W.  R.  Addicks,  president  of  the 
American  Gas  Institute,  invited  the 
council  to  appoint  a  representative  of 
the  society  to  a  committee  which  is  to 
arrange  for  a  gas  congress  in  San 
Francisco  during  the  Exposition  in  1915. 


TRANSACTIONS    I.    E.    S.  —  PART    I 


The  president  was  authorized  to  appoint 
this   representative. 

The  1912  annual  report  of  the  gen- 
eral secretary  was  received.  This  re- 
port with  certain  modifications  was 
made  the  report  of  the  council  to  the 
society;  it  appears  in  full  elsewhere  in 
this  issue  of  the  Transactions. 
Those  present  at  the  meeting  were : 
V.  R.  Lansingh.  president;  T.  D. 
Israel.  A.  J.  Marshall,  P.  W.  Cobb, 
George  S.  Barrows,  H.  E.  Ives,  J.  \V. 
Cowles.  R.  C.  Ware.  W.  J.  Serrill.  J.  T. 
Maxwell.  L.  B.  Marks,  C.  J.  Russell, 
E.  P.  Hyde.  C.  H.  Sharp,  E.  B.  Rosa, 
A.  E.  Kennelly,  C.  O.  Bond.  Norman 
Macbeth,  and  Preston  S  Millar,  general 
secretary.  Mr.  W.  R.  Addicks.  presi- 
dent of  the  American  Gas  Institute,  was 
present  upon  invitation. 

SPECIAL   COUNCIL    MEETING 

A  special  meeting  of  the  council,  the 
first  meeting  of  the  new  administration. 
was  held  in  the  Engineers'  Club,  New 
York  City,  January  10,  1913.  Those  in 
attendance  were : 

Preston  S.  Millar,  president ;  J.  D. 
Israel,  general  secretary;  C.  J.  Russell. 
George  S.  Barrows.  C.  H.  Sharp,  E.  B. 
Rosa,  W.  J.  Serrill.  C.  O.  Bond.  P.  W. 
Cobb,  R.  C.  Ware.  L.  B.  Marks.  E.  P. 
Hyde.  H.  E.  Ives.  J.  T.  Maxwell,  J.  W. 
Cowles,  Norman  Macbeth,  and  V.  R. 
Lansingh. 

After  a  discussion  of  the  question  of 
sustaining  membership,  it  was  resolved 
that  the  dues  of  sustaining  members  be 
left  to  the  discretion  of  the  committee 
(committee  to  be  appointed  by  the 
president)  having  this  matter  in  charge, 
that  the  amount  of  such  dues  be  not 
published  in  the  by-laws,  and  that  the 
rules  and  regulations  which  may  be  pro- 
posed by  the  committee  be  first  approved 
by  the  council  or  its  executive  com- 
mittee. 


A  list  of  changes  in  the  by-laws,  most 
of  which  were  necessitated  by  the  recent 
constitutional  change  in  the  fiscal  year 
of  the  society,  was  read  a  second  time 
and  adopted. 

President  Millar  announced  his  ap- 
pointments to  various  standing  and 
temporary  committees.  The  appoint- 
ments were  approved.  It  was  under- 
stood that  additional  appointments  to 
these  committees  would  be  made  later. 
A  list  of  committees  and  the  personnel 
of  each  appears  in  the  front  part  of  this 
issue  of  the  Transactions. 

An  appropriation  of  $100  was  author- 
ized to  cover  the  cost  of  printing  a 
prospectus  on  the  society  and  its  work. 

The  president  was  authorized  to 
appoint  a  committee  of  five  to  investi- 
gate, and  report  to  the  February  council 
meeting,  the  matter  of  appointing  rep- 
resentatives or  local  secretaries  in  cities 
not  having  sections  of  the  society. 

It  was  decided  to  hold  the  regular 
meetings  of  the  council  during  the  pres- 
ent administration  in  the  morning  of 
the  second  Friday  of  each  month — 
except,  of  course,  during  the  months  of 
July,  August  and  September,  when  no 
regular  meetings  are  held. 


New  Members. 

At  a  meeting  of  the  council  held  in 
New  York,  January  10,  the  following 
applicants  were  elected  members  of  the 
society : 

Arrighi,  Roswell. 

Agent,  The  New  York  Edison  Com- 
pany, 124  West  42nd  Street,  New 
York. 

Baker,  Cyrus  Rex  ford. 

Incandescent  Lamp  Specialist.  Gen- 
eral Electric  Co.,  30  Church  Street, 
New  York. 


TRANSACTIONS   I.    E.    S. — PART   I 


BlERMAN,   CHAS. 

Telephone  Engineer,  Wisconsin 
Telephone  Co.,  183  gth  Street,  Mil- 
waukee, Wis. 

Brown,  Melvin   P. 

Lighting  Inspector,  Dept.  Water 
Supply,  Gas  &  Electricity,  13-21 
Park  Row,  New  York. 

Campbell,  O.  M. 

Sales  Engineer,  National  X-Ray 
Reflector  Co.,  235  W.  Jackson 
Boulevard,  Chicago,  111. 

Cole,  Chas.  M. 

Illuminating  Engineer,  Wheeler  Re- 
flector Co.,  156  Pearl  Street,  Bos- 
ton, Mass. 

Conner,  George  C. 

Engineer,  National  Electric  Lamp 
Association,  441 1  Hough  Avenue, 
N.  E.,  Cleveland,  Ohio. 

Dalton,  Parker  C. 

Salesman,  Philadelphia  Electric  Co., 
1000  Chestnut  Street,  Philadelphia, 
Pa. 

Duane,  Dr.  Alexander. 

129  East  37th  Street,  New  York. 

Ferguson,  Joseph  Simpson. 

Student,  Philadelphia  Trades 
School,  12th  and  Locust  Streets, 
Philadelphia,   Pa. 

Galavan,  Edward. 

Sales  Engineer,  National  X-Ray 
Reflector  Co.,  235  W.  Jackson 
Boulevard,  Chicago,  111. 

Keane,  H.  P. 

Sales  Engineer,  National  X-Ray 
Reflector  Co.,  235  W.  Jackson 
Boulevard,  Chicago,  111. 

Knight,  J.  Harmer. 

Draftsman,  Philadelphia  Electric 
Co.,  10th  and  Chestnut  Streets, 
Philadelphia,    Pa. 

La  Belle,  John  N. 

Supervising  Engineer,  National 
X-Ray  Reflector  Co.,  235  W.  Jack- 
son Boulevard,  Chicago,  111. 


Lewis,  Dr.  F.  Park. 

454  Franklin  Street,  Buffalo,  N.  Y. 

Martin,  W.  G. 

Engineer,  National  X-Ray  Reflector 
Co.,  235  W.  Jackson  Boulevard, 
Chicago,  111. 

Mass,  Herbert  C. 

Illuminating  Engineer,  436  Henry 
Bldg.,  Seattle,  Wash. 

Maxwell,  C.  M. 

Electrical  Draughtsman,  R.  D.  Kim- 
ball Co.,  15  West  38th  Street,  New 
York. 

McKinnie,  E.  C. 

Engineer,  National  X-Ray  Reflec- 
tor Co.,  235  W.  Jackson  Boulevard, 
Chicago,  111. 

Parrott,  Robert. 

Sales  Engineer,  General  Electric 
Co.,  30  Church   Street,   New  York. 

Selleck,  John  K. 

Engineer,  National  X-Ray  Reflec- 
tor Co.,  235  W.  Jackson  Boulevard, 
Chicago,  111. 

Seymour,  F.  W. 

Lighting  Inspector,  Dept.  Water 
Supply,  Gas  &  Electricity,  13-21 
Park  Row,  New  York. 

States,  Wilmer  M. 

Salesman,  General  Electric  Co., 
Edison  Lamp  Works,  Harrison, 
N.J. 

Sullivan.  J.  B. 

General  Electric  Company,  Foreign 
Department,  Sarmiento  531,  Buenos 
Aires,  Arg. 

Tolman,  W.  H. 

Director,  Museum  of  Safety,  29 
West  39th  Street,  New  York. 


Section  Activities. 

CHICAGO    SECTION 

A  regular  meeting  of  the  Chicago 
section  was  held  in  the  auditorium  of 
the  Western  Society  of  Engineers, 
Chicago,    January    15,    1913.      Sixty-five 


TRANSACTIONS    I.    E.    S. — PART    I 


members  and  guests  were  present.  Mr. 
T.  H.  Aldrich  of  the  National  X-Ray 
Reflector  Company  and  Mr.  J.  P.  Malia, 
chief  electrician  of  Armour  &  Com- 
pany, presented  a  paper  entitled  "Indi- 
rect Illumination  of  General  Offices." 
The  paper  was  for  the  most  part  a 
detailed  description  of  the  lighting  in- 
stallation in  the  general  offices  of 
Armour   &   Company. 

The  following  program  of  meetings 
has  been  arranged : 

February  22 — A  joint  meeting  of 
engineers,  architects  and  ophthalmolo- 
gists, several  organizations  participating, 
in  Milwaukee,  Wis. 

March  19 — Announcement  will  be 
made  later. 

NEW    ENGLAND    SECTION 

A  regular  meeting  of  the  New  Eng- 
land section  was  held  in  the  auditorium 
of  the  Edison  Electric  Illuminating 
Company,  January  21,  1913.  Two  papers 
were  read :  one  on  "Commercial  Lenses" 
by  Dr.  H.  P.  Gage  of  the  Corning  Glass 
Company,  Corning,  N.  Y. ;  the  other, 
"Problems  of  Lighthouse  Service  and 
How  They  Met"  by  Dr.  Raymond 
Haskell  of  the  United  States  Lighthouse 
Service.  Both  papers  were  illustrated 
by  lantern  slides  and  were  very  interest- 
ing. Preceding  the  meeting  a  dinner 
was  held  at  "The  Georgian,"  at  which 
plans  for  the  general  welfare  of  the 
section  were  discussed. 

The  following  program,  subject  to 
change,  has  been  announced : 

February  17 — Joint  meeting  with  the 
Boston  section  of  the  American  Insti- 
tute of  Electrical  Engineers.  Papers : 
"Ornamental  Magnetite  Arc  Lamps"  by 
C.  A.  B.  Halvorson  of  the  General  Elec- 
tric Company,  West  Lynn,  Mass. ;  "The 
Enclosed  Flame  Arc  Lamp"  by  W.  A. 
Darrah  of  the  Westinghouse  Electric  & 
Manufacturing    Company,     East    Pitts- 


burgh,    Pa. ;     "Mercury-Vapor    Lamps" 
by  P.  H.  Thomas,  New  York. 

March  18 — A  demonstration  of  inte- 
rior lighting  effects,  by  Preston  S. 
Millar. 

NEW    YORK    SECTION 

The  New  York  section  held  a  joint 
meeting  with  the  National  Commercial 
Gas  Association  in  the  United  Engineer- 
ing Societies'  Building,  January  9,  1913. 
Two  papers — "The  Lighting  of  Taft 
Hall  in  the  Auditorium  Armory"  (At- 
lanta, Ga.)  by  Robert  F.  Pierce  of  the 
Welsbach  Company,  and  "The  Lighting 
of  the  Exhibition  Hall,  Auditorium 
Armory"  (Atlanta,  Ga.),  by  J.  M.  Coles, 
were  presented.  About  175  attended 
the  meeting.  Preceding  the  meeting 
there  was  an  informal  dinner  at  Keene's 
Chop  House. 

The  program  of  meetings  for  the  re- 
mainder of  the  season  is  as  follows : 

February  7 — A  joint  meeting  with  the 
Municipal  Art  Society  at  the  New  York 
Arts  Club. 

March  13 — Joint  meeting  with  the 
American  Society  of  Mechanical  Engi- 
neers in  the  United  Engineering  So- 
cieties Building,  29  West  39th  Street, 
New  York.  Mr.  Ward  Harrison  of  the 
National  Electric  Lamp  Association  will 
present  a  paper  on  "Industrial  Light- 
ing." 

April  8 — This  meeting  will  probably 
be  held  in  the  United  Engineering 
Societies  Building.  Mr.  M.  Luckiesh  of 
the  National  Electric  Lamp  Association 
will  present  a  paper  on  "Light  and  Art." 
A  paper  on  "Phosphorescence  and 
Fluorescence"  is  also  scheduled.  This 
meeting  should  be  an  unusually  inter- 
esting one. 

May  8 — A  talk  on  theater  lighting  by 
Mr.  Bassett  Jones,  Jr.,  at  the  Clymer 
Street  Theater,  Brooklyn.  During  the 
past   year    Mr.    Jones   has    conducted    a 


8 


TRANSACTIONS    I.    E.    S. — PART    I 


great  deal  of  experimental  work  in 
theater  illumination  particularly  in  the 
production  of  stage  effects.  The  mem- 
bers of  the  New  York  chapter  of  the 
American  Institute  of  Architects  will  be 
invited  to  attend  this  meeting.  Admis- 
sion will  be  by  card. 

June  8 — It  is  planned  to.  have  a  joint 
meeting  and  outing  of  all  the  engineer- 
ing societies  in  New  York. 

PHILADELPHIA   SECTION 

A  joint  meeting  with  the  Philadelphia 
section  of  the  American  Institute  of 
Electrical  Engineers  and  the  Philadel- 
phia Electric  Company  section  of  the 
National  Electric  Light  Association  was 
held  at  the  Engineers'  Club,  1317  Spruce 
St.,  Jan.  13.  Short  talks  on  the  subject 
of  "Modern  Illumination"  were  given  by 
Prof.  George  A.  Hoadley  of  Swarth- 
more  College,  Prof.  Arthur  J.  Rowland 
of  Drexel  Institute,  and  Joseph  D. 
Israel  of  the  Philadelphia  Electric  Com- 
pany. Dr.  G.  S.  Crampton,  W.  E.  Rob- 
ertson, H.  Calvert,  B.  Frank  Day,  R.  F. 
Pierce,  and  G.  H.  Swanfeld  also  gave 
brief  talks  on  various  phases  of  the 
subject  of  illumination  Mr.  R.  B.  Ely 
exhibited  a  number  of  modern  illumi- 
nants. 

A  number  of  members  of  the  Phila- 
delphia section  attended  the  meeting  of 
the  Franklin  Institute  and  the  Philadel- 
phia Electric  Company  section  of  the 
National  Electric  Light  Association  in 
the  auditorium  of  the  Institute,  15  South 
Seventh  Street,  Thursday  evening,  Jan- 
uary 30.  At  that  meeting  Dr.  E.  P. 
Hyde  presented  a  paper  on  "The  Phys- 
ical Laboratory  of  the  National  Electric 
Lamp  Association."  Dr.  Hyde's  paper 
was  supplemented  by  a  series  of  lantern 
slides  showing  the  new  buildings  of  the 
association  which  occupy  a  forty-acre 
plot  in  Cleveland,  which  is  to  be  known 


as  Nela  Park.  Dr.  Hyde  stated  that 
when  the  buildings  are  completed  there 
will  be  available  excellent  facilities  for 
the  conduct  of  the  scientific  problems 
of  the  lighting  industry. 

The  following  meetings  have  been 
scheduled: 

February  21— A  demonstration  of 
interior  lighting  effects  by  Preston  S. 
Millar. 

During  the  week  of  March  23  a  meet- 
ing will  be  held  in  the  New  Century 
Drawing  Rooms.  Mr.  M.  Luckiesh  of 
the  physical  laboratory  of  the  National 
Electric  Lamp  Association  will  present 
a  paper  on  "Light  and  Art."  Notices  of 
this  meeting  will  be  issued  shortly. 

The  dates  and  papers  for  subsequent 
meetings  will  be  announced  later. 

PITTSBURGH     SECTION 

Mr.  H.  W.  Shalling  read  an  interest- 
ing paper  on  "Department  Store  Light- 
ing" at  a  meeting  of  the  Pittsburgh  sec- 
tion, January  24.  The  paper  and  its 
attending  discussion  appears  in  this 
issue  of  the  Transactions.  The  mem- 
bers in  attendance  were  the  guests  of 
McCreery  &  Company,  in  whose  store 
the  meeting  was  held.  At  the  conclu- 
sion of  the  meeting  a  resolution  of 
thanks  to  McCreery  &  Company  was 
adopted. 

The  program  of  meetings  for  the  rest 
of  the  season  is  as  follows : 

February — "Gas  Lighting"  by  S.  B. 
Stewart. 

March— "Moving  Picture  Lanterns 
from  the  Central  Station  Point  of 
View"   by  J.   F.   Martin. 

April — "Railroad  Car  Lighting"  by 
J.  L.  Minick. 

May— "Physiological  Aspects  of  Il- 
lumination" by  W.  E.  Reed. 

June — Announcement  will  be  ~  made 
later. 


TRANSACTIONS    I.    E.    S.—  PART    I 


9 


Annual  Meeting. 

Sixty-nine  members  and  guests  were 
present  at  the  annual  meeting  which 
was  held  in  the  Aldine  Club,  Fifth 
Avenue  and  23rd  Street,  New  York, 
January  io,  1913.  A  dinner  preceded  the 
meeting.  Brief  addresses  on  the  society 
and  various  phases  of  its  work  were 
made  by  Dr.  A.  E.  Kennelly  of  Harvard 
University,  Mr.  T.  C.  Martin,  secretary 
of  the  National  Electric  Lamp  Associa- 
tion, Dr.  W.  H.  Tolman,  director  of  the 
American  Museum  of  Safety,  Dr.  C.  H. 
Sharp  of  the  Electrical  Testing  Labora- 
tories, Mr.  L.  B.  Marks,  first  president 
of  the  society,  Mr.  V.  R.  Lansingh,  the 
retiring  president;  Mr.  W.  R.  Addicks, 
president  of  the  American  Gas  Institute, 
and  Mr.  J.  W.  Lieb,  Jr.,  vice-president 
of  the  New  York  Edison  Company. 

During  the  meeting  it  was  announced 
that  the  following  officers  had  been 
elected  at  the  previous  election :  Preston 
S.Millar,  president;  vice-presidents, Wm. 
J.  Serrill,  J.  W.  Cowles,  J.  R.  Cravath, 
H.  S.  Evans;  general  secretary  Joseph 
D.  Israel;  L.  B.  Marks,  treasurer;  and 
C.  O.  Bond,  P.  W.  Cobb  and  W.  Cullen 
Morris,  directors.  The  announcements 
were  received  with  applause.  It  was 
also  reported  that  the  constitutional 
amendments  which  had  been  submitted 
at  the  election  had  been  adopted. 

Retiring  President  Lansingh  then  in- 
troduced President  Millar,  who  deliv- 
ered his  inaugural  address  on  the 
progress  and  functions  of  the  society. 
The  address  is  printed  elsewhere  in 
this  issue  of  the  Transactions.  A 
brief  address  was  also  delivered  by  Mr. 
Joseph  D.  Israel,  the  newly  elected  gen- 
eral secretary. 


The  report  of  the  council  covering  the 
work  of  past  year  was  presented  in  an 
abstract  form.  The  full  report,  the  re- 
port of  the  general  secretary,  is  printed 
in  this  number. 


A  New  By-Law. 

The  following  by-law  which  outlines 
the  procedure  of  section  nominating 
committees  was  adopted  at  a  meeting  of 
the  council,  January  10,   1913 : 

The  procedure  in  nominating  and  electing 
section  officers  shall  be  as  follows,  except  when 
other  procedure  shall  be  authorized  by  the 
Council. 

A  section  nominating  committee  shall  be  ap- 
pointed by  the  Section  Board  of  Managers  each 
year.  The  appointment  shall  be  reported  to  the 
General  Secretary.  This  committee  shall  con- 
sist of  five  members  of  whom  at  least  two  shall 
be  past  officers  of  the  Section  or  members  of  the 
Council.  Not  later  that  March  15  of  each  year, 
the  General  Secretary  shall  notify  the  chairman 
of  the  committee  that  it  is  the  committee's  duty 
to  prepare  a  nomination  ticket  containing  the 
names  of  those  whom  they  deem  best  suited  for 
the  section  offices  to  be  filled  at  the  ensuing  annual 
election.  The  report  of  the  committee  shall  be 
prepared  in  duplicate,  one  copy  shall  be  sub- 
mitted to  the  chairman  of  the  section  and  the 
other  copy  shall  be  delivered  to  the  General 
Secretary  not  later  than  April  15.  The  ticket 
thus  prepared  by  the  committee  on  nomination 
shall  be  printed  and  forwarded  to  all  section 
members  not  later  than  May  5,  in  connection 
with  the  ballots  for  election  of  general  officers. 
The  election  of  section  officers  in  other  respects 
shall  be  carried  out  in  a  manner  similar  to  that 
prescribed  for  the  election  of  general  officers, 
save  that  a  copy  of  the  report  of  the  Committee 
of  Tellers  on  the  results  of  the  section  election 
shall  be  mailed  as  soon  as  prepared,  to  the  chair- 
man of  the  section  and  to  the  chairman-elect. 

This  by-law  will  standardize  and 
facilitate  the  procedure  in  electing  sec- 
tion officers.  It  will  also  eliminate  con- 
siderable work  and  expense  in  connec- 
tion  with  the  elections. 


10  TRANSACTIONS   I.    E.    S. — PART  I 

Annual  Report  of  the  Finance  Committee  for  the  Fiscal  Year  1912. 


To  the  Council  of  the  Illuminating  Engineering  Society: 

In  accordance  with  the  provisions  of  the  constitution  of  the  society,  the 
Finance  Committee  exercised  direct  supervision  over  the  financial  affairs  of  the 
society. 

The  committee  held  a  meeting  each  month  except  during  July,  August  and 
September,  examined  and  approved  all  bills  paid  by  the  society,  and  presented  a 
written  report  at  each  meeting  of  the  council. 

The  financial  condition  of  the  society  as  of  Dec.  31,  1912,  is  given  in  the  sub- 
joined statement  of  Messrs.  William  J.  Struss  &  Co.,  certified  public  accountants, 
who  were  employed  by  authorization  of  the  council  to  audit  the  books  and  accounts 
of  the  society. 

The  auditor's  report  shows  a  deficit  of  $640.07  for  the  year  and  an  impairment 
of  surplus  amounting  to  $1,435.63  since  Jan.  1,  1912. 

The  membership  of  the  society  at  the  close  of  the  year  was  1,325.  At  one 
time  during  the  year  the  membership  reached  1,470.  Based  on  an  average  active 
membership  of  1,350  for  the  year,  the  expenses  per  member  were  $7.36.  The 
income,  other  than  that  obtained  from  membership  dues,  was  derived  chiefly  from 
the  proceeds  of  advertising  and  miscellaneous  sales  of  the  Transactions  of  the 
society. 

Early  in  1912  the  committee  reported  to  the  council  that  the  society  was  likely 
to  face  a  considerable  deficit  at  the  close  of  the  year,  and  recommended  that  imme- 
diate steps  be  taken  to  secure  a  larger  income  from  the  membership.  The  council 
appointed  a  special  committee  on  "Financial  Policy"  and  subsequently  a  committee 
on  "Revenue,"  to  consider  ways  and  means  of  placing  the  society  on  a  sound 
financial  basis  and,  as  a  result  of  protracted  discussion  of  the  subject,  the  plan 
of  "Sustaining  Membership"  (now  forming  part  of  the  constitution)  was  finally 
evolved.  If  this  plan  works  out  as  well  as  expected,  the  income  derived  from 
the  membership  at  large  will  be  sufficient  not  only  to  defray  the  ordinary  running 
expenses  of  the  society,  but  also  to  meet  increased  expenses  due  to  expansion  of 
the  activities  of  the  society. 


Respectfully  submitted, 


A.  A.  Pope, 

A.  S.  McAllister, 

L.  B.  Marks,  Chairman. 


TRANSACTIONS    I.    E.    S. — PART    I  11 

STATEMENT  OF  THE  AUDITORS. 
Exhibit  "A." — Balance  Sheet,  December  31,  1912. 

ASSETS. 
Cash- 
On  hand  and  in  bank $1,918.23 

191 3 — New  York  Section  expenses  paid   1912 29.25 

Accounts   Receivable — 

1912  dues    $  12.50 

Miscellaneous  accounts    510.01 

Initiation    fees    10.00 

1912    advertising    402.38 

Total     934.89 

Property  Accounts — 

Furniture   and   fixtures    630.21 

Less  depreciation — 15  per  cent 94-53 

Net    535.68 

Badges  on  hand    (29) 84.00 

Total    :.      619.68 

Investments — 

Northern  Pacific  and  Great  Northern  Railway  Bonds — $2,000 1,920.00 

Total    $5,422.05 

liabilities. 

Accounts   payable    $   803.57 

Advance  dues  2,060.00 

Advance  fees   5.00 

December  expenses  estimated — Exhibit  "B,"  Schedule  No.  1 1,080.00 

Advance  advertising   19-54 

Surplus — Exhibit  "A,"  Schedule  No.  1 1,453.94 

Total    $5,422.05 

Exhibit  "A,"  Schedule  No.  i — Surplus  Account,  Dec.  31,  1912. 

Surplus — January  1,  1912  $2,889.57 

Duplicate  charge  in  191 1 9.98 

Back  dues    10.00 

2.009.55 


12  TRANSACTIONS   I.    E.    S.  — PART   I 

1911   New  York  Section  expenses $  62.10 

191 1    General    Office    60.30 

191 1  Transactions 287.66 

191 1    Philadelphia   Section  expenses    28.85 

191 1    Chicago   Section    expenses. 56.49 

Dropped  (various)   in  default  of  fees 74.00 

1911   Election  expenses    (part) 163.85 

191 1  Membership   Committee    (part) 82.29 

Deficit  for  year  1912   (see  Exhibit  "B") 640.07 

I45S-6I 

$1,453-94 
Exhibit  "B" — Statement  of  Earnings  and  Expenses  for  the  Year 
Ended  December  31,  1912. 
earnings. 

Members'   dues    $6,872.19 

Advertising    1,356.22 

Miscellaneous   sales   of   Transactions 586.78 

Initiation    fees    375-00 

Interest  on  bonds   80.00 

Profit  on  badges  sold 21 .00 

Members'    certificates    8.00 

Total $9,299.19 

expenses. 

Transactions    $2,253.29 

December  expenses  estimated  (Exhibit  "B" — Schedule  No.  1)....   1,080.00 

General  Office  (Exhibit  "B" — Schedule  No.  2) 4,057.34 

New   York    Section    389.83 

Chicago    Section    251.17 

New   England    Section    208.43 

Pittsburgh  Section    148.31 

Philadelphia    Section    291.32 

Committee  on  Illumination  Primer 463.15 

1912  Convention  Committee    420.14 

1912  Election  expense   1 14.12 

Depreciation — furniture   and   fixtures    94-53 

Papers    Committee    44-07 

Committee   on    Reciprocal    Relations 4.25 

Committee  on  Nomenclature  and  Standards i5-7o 

Joint  meetings  with  other  societies 27.43 

Annual   Meeting    32-75 

Authors'   advance  copies 5.38 

Treasurer's    expense    16.44 

President's   expense    9.25 

Exchange  on  checks    12.36 

Total     9,939.26 

Excess   of   Expenses   over   Earnings $640.07 


TRANSACTIONS    I.    E.    S. — PART    I  13 

Exhibit  "B,"  Schedule  No.  i — December   (1912)   Expenses,  Estimated. 

Primer     $680.00 

Transactions     250.00 

Chicago    Section    10.00 

Philadelphia    Section    20.00 

Pittsburgh   Section    5.00 

New    England    Section    5.00 

General    Office    (part)     50.00 

Miscellaneous     60.00 


$1,080.00 


Exhibit  "B,"  Schedule  No.  2 — Analysis  of  General  Office  Account 
for  the  12  Months  Ended  Dec.  31,  1912. 

Salaries  :    Assistant  Secretary  and  Stenographer $2,210.68 

Rent     666.00 

Postage     298.57 

Telephone  and   telegraph    152.98 

Printing,   stationery,   etc 332.37 

Miscellaneous     396.74 

$4,057-34 


TRANSACTIONS 

OF  THE 

Illuminating 
Engineering  Society 

JANUARY,  1913 
PART  II 

Papers,   Discussions  and  Reports 


[  JANUARY,  1913  ] 

CONTENTS  -  PART  II 


Inaugural  Address  of  President  P.  S.  Millar i 

1912  Report  of  the  General  Secretary 6 

Department  Store  Lighting.     By  H.  W.  Shalling  17 

Tests  for  the  Efficiency  of  the  Eye  Under  Different  Systems 
of  Illumination  and  a  Preliminary  Study  of  the  Causes  of 
Discomfort.     By  C.  E.  Ferree 49 


INAUGURAL  ADDRESS  OF  PRESIDENT  MILLAR.* 


Fellow  Members  of  the  Illuminating  Engineering  Society: 

We  who  are  members  feel  sure  that  this  society  is  destined 
to  fulfill  an  important  purpose.  The  problem  which  it  must 
solve  is  tremendously  difficult  and  almost  unique  in  its  com- 
plexity. The  goal  has  been  pointed  out  and  described  by  Past- 
President  Hyde  in  these  words : 

The  goal  of  illuminating  engineering  will  have  been  attained  when  as  a 
result  of  the  concomitant  development  of  its  component  elements,  it  will  be 
possible  in  every  case  presented,  to  design  a  lighting  installation  which  will 
be  efficient,  effective,  artistic  ;  which  will  produce  an  illumination  correct 
in  quantity  and  quality  properly  balanced  as  to  high-light  and  shadow,  rest- 
ful to  the  eye  and  harmonious  with  the  form  and  color  schemes  involved  ; 
which  will  stand  the  rigorous  test  of  logical  analysis,  and  will  appeal  to  the 
highly  developed  sense  of  beauty.  The  goal  of  illuminating  engineering  is 
the  attainment  to  the  ideal  application  of  perfect  knowledge. 

At  the  time  when  the  society  was  organized  we  were  far  from 
the  goal.  Many  of  us  did  not  know  what  or  where  the  goal  is. 
Our  lighting  practise  in  general  was  execrable.  To-day  we  are 
well  started  on  our  way.  While  the  goal  is  still  so  far  away 
that  we  cannot  afford  to  interrupt  our  journey  to  celebrate 
progress,  yet  we  may  pause  a  moment  to  look  backward  and 
contemplate  with  gratification  the  improvement  which  has  taken 
place  in  lighting  practise  and  the  advance  made  in  our  knowledge 
of  lighting  principles.  From  such  contemplation  we  may  derive 
inspiration  for  the  great  journey  still  before  us. 

The  history  of  the  society  covers  seven  years.  They  have 
been  busy  years.  Much  has  been  accomplished  though  that  much 
seems  little  when  compared  with  the  great  work  remaining  to  be 
accomplished. 

The  first  year  was  devoted  to  establishing  the  society,  bringing 
into  its  membership  those  connected  with  various  industries  and 
professions  concerned,  and  securing  papers  on  some  of  the  many 
aspects  of  the  subject  of  lighting,  in  order  to  introduce  the  society 
and  start  it  on  its  way.  Through  the  Herculean  efforts  of 
President  Marks  the  year  closed  with  more  than  800  members, 
surely  a  phenomenal  record. 

*  Delivered  at  the  annual  meeting  of  the  Illuminating  Engineering  Society,  New 
York,  January  10,  1913. 


2  TRANSACTIONS  I.  E.  S. — PART  II 

The  second  year,  Dr.  Sharp's  administration,  was  devoted  to 
the  establishment  of  sections,  the  extension  of  the  society's 
influence  throughout  various  cities  and  the  perfection  of  its 
organization.  Many  vexing  problems  of  an  intra-society  character 
were  met  and  disposed  of.  The  foundation  was  laid  for  the 
society's  later  work. 

In  the  third  year,  Dr.  Bell's  administration,  the  society  may 
be  said  to  have  found  itself  as  an  organization.  The  various 
officers  attended  to  their  functions  and  a  display  of  team  work 
made  it  evident  that  this  is  no  one-man  enterprise,  but  a  well 
organized  society  with  many  actively  interested  members. 

During  the  fourth  year,  under  the  administration  of  President 
Gartley,  internal  affairs  were  further  strengthened  and  the  pur- 
pose and  work  of  the  society  was  brought  before  the  gas  indus- 
try in  a  way  to  enlist  more  support  than  had  ever  before  been 
accorded. 

In  the  fifth  year  our  organization  was  found  to  be  strong 
enough  to  attempt  something  beyond  the  continued  development 
of  its  internal  activities  and  the  lecture  course  on  illuminating 
engineering  was  conceived  by  Dr.  Hyde  and  carried  through 
with  notable  success.  Thus  was  the  scope  of  illuminating  engi- 
neering defined,  a  clearer  conception  of  its  various  elements 
placed  before  the  society  and  the  public,  and  the  ground-work 
laid  for  university  educational  courses  in  illumination. 

The  sixth  year  witnessed  an  extension  of  work  to  include  the 
education  of  the  public  in  lighting  fundamentals.  Work  was 
started  upon  a  primer  of  illumination.  At  the  same  time  another 
step  was  taken  toward  the  establishment  of  the  profession  of 
illuminating  engineering  through  Dr.  Kennelly's  scholarly  in- 
augural address. 

During  President  Lansingh's  administration  the  illumination 
primer  has  been  completed  and  its  dissemination  begun.  The 
society  has  given  additional  evidence  of  realization  of  respon- 
sibility to  humanity  by  the  appointment  of  a  Committee  on 
Glare  from  Reflecting  Surfaces.  It  has  sought  to  discharge  its 
civic  duty  through  the  activity  of  a  Committee  on  Factory  Light- 
ing Legislation  which  has  co-operated  with  the  New  York  State 
Labor  Commission  in  devising  constructive  but  safe  and  sane 
legislation  on  industrial  illumination. 


INAUGURAL  ADDRESS   OF   PRESIDENT    MILEAR  3 

Thus  we  find  that  the  first  four  years  of  the  society's  work 
were  devoted  to  perfecting  its  organization,  improving  its  inter- 
nal mechanism,  and  developing  its  function  as  a  forum  for  dis- 
cussion. With  the  fifth  year  came  activity  looking  toward  the 
creation  of  a  profession  of  illuminating  engineering.  This  was 
followed  quickly  by  efforts  toward  public  education.  A  per- 
fectly logical  course  of  action,  involving  first  preparation  for 
the  task,  and  after  at  least  some  material  progress  has  been 
made  within  the  society  as  preparation,  the  application  outside 
the  society  of  the  knowledge  acquired  in  an  effort  to  benefit  the 
public  and  establish  the  profession. 

As  I  see  it  therefore  we  have  arrived  at  a  stage  of  develop- 
ment where  our  functions  are  defined  not  alone  by  declaration 
but  by  actions  as  well.     These  functions  are  three  in  number. 

First,  to  serve  as  a  forum  for  the  presentation  and  discussion 
of  technical  questions  pertaining  to  light  and  illumination, 
thereby  promoting  the  advance  of  knowledge  and  informing  the 
membership  of  knowledge  acquired ; 

Second,  to  improve  lighting  practise  through  the  formulation 
and  application  of  principles  of  good  illumination  and  through 
the  education  of  lighting  practitioners  and  the  public  at  large  in 
matters  of  illumination ; 

Third,  ultimately,  perhaps,  to  establish  a  professional  basis  for 
illuminating  engineering. 

In  regard  to  the,  third  function,  I  believe  that  we  are  all  in 
accord  in  a  desire  to  promote  developments  which  will  contribute 
toward  the  establishment  of  a  professional  basis  for  illumi- 
nating engineering;  but  it  does  not  appear  to  me  that  the  time 
is  ripe  for  determined  effort  in  that  direction.  I  believe  that 
without  neglecting  this  aspect  of  our  problem  we  may  direct  our 
greatest  energies  more  profitably  toward  accomplishment  in  other 
directions.  Give  education,  that  great  panacea,  some  further 
opportunity  to  create  the  demand ;  advance  the  boundaries  of 
knowledge  of  illumination ;  devote  efforts  to  earning  greatest 
respect  for  the  society ;  and  all  in  good  time  the  need  for  such 
a  specialty  will  be  generally  recognized  and  it  will  be  found  prac- 
ticable to  establish  the  desired  professional  basis. 

As    a    forum,   the    society   is   concerned   with    the   science   of 


4  TRANSACTIONS  I.  E.  S. PART  II 

illumination.  Knowledge  must  precede  application.  It  was 
inevitable  therefore  that  this  part  of  the  work  should  have  feat- 
ured the  earlier  years.  Thus  the  mathematics  of  illumination, 
well  grounded  by  Dr.  Sharp  in  his  presidential  address  in  1907, 
has  been  thoroughly  developed  and  has  received  a  full  share  of 
attention  in  our  deliberations.  The  measuremnet  of  light  as  an 
element  of  fundamental  importance  has  been  widely  discussed, 
though  much  remains  for  future  development.  Papers  present- 
ing experience  of  lighting  practitioners  have  naturally  predomi- 
nated and  in  the  aggregate  have  added  much  to  our  store  of 
knowledge.  These  are  among  the  many  phases  of  the  science 
of  illumination  which  have  received  treatment  commensurate 
with  their  importance.  But  there  are  some  phases  which,  con- 
trary to  the  desire  and  in  spite  of  the  effort  of  the  successive 
administrations,  have  not  been  adequately  discussed.  Among 
these  are  the  principles  of  architecture  and  decoration  as  con- 
stituting requirements  for  lighting  design.  Our  science  is  defi- 
cient in  this  respect,  and  there  is  no  more  important  need  to  be 
met  than  the  supply  of  knowledge  to  meet  this  deficiency. 

The  art  of  illumination,  in  the  sense  of  application  of  knowl- 
edge, has  been  developed  rapidly  with  the  advancement  of  the 
science.  Newly  designed  installations,  and  new  lighting  equip- 
ments are  greatly  in  advance  of  those  of  eight  years  ago.  But 
the  many  installations  which  transgress  principles  of  hygiene, 
esthetics  or  economy  are  every-day  evidence  that  relatively  little 
has  been  accomplished.  And  the  remedy  for  this  we  believe  is 
education — education  of  our  members  and  others  who  have  to 
do  with  the  installation  of  lighting  equipment ;  education  of 
members  of  organizations  who  may  become  interested  in  improv- 
ing lighting  conditions;  education  of  students  in  our  universities 
and  schools ;  education  of  the  public  at  large.  Many  of  us  find 
ourselves  peculiarly  sensitive  to  untoward  effects  of  bad  light- 
ing because  we  have  studied  illumination  and  recognize  bad  con- 
ditions to  which  we  formerly  were  unconscious.  Our  problem 
is  to  educate  the  American  people  to  be  similarly  impressed  by 
bad  lighting.  When  that  is  done  most  of  the  very  bad  lighting 
will  be  improved.  To  succeed,  the  society  must  address  itself 
to  the  task  of  furthering  this  educational  work  upon  which 
such  an  excellent  start  has  been  made. 


INAUGURAL   ADDRESS   OF   PRESIDENT   MIELAR  5 

The  society  this  year  is  committed  to  a  policy  of  expansion. 
We  are  to  seek  to  extend  our  influence  in  a  number  of  ways. 

The  establishment  of  local  representatives  in  cities  where  there 
is  no  section  is  expected  to  result  in  a  more  general  knowledge  of 
our  purpose  and  work,  and  to  promote  the  application  of  prin- 
ciples of  good  lighting.  In  arousing  and  maintaining  local  interest 
in  lighting  matters,  this  is  expected  to  extend  greatly  the  society's 
influence  for  good. 

Sustaining  membership  provides  a  means  whereby  lighting  com- 
panies, manufacturers  and  others  who  recognize  the  value  of  the 
society's  work  to  them  may  give  a  limited  amount  of  support  in 
application  of  that  work.  Of  equal  importance,  it  gives  the 
society  an  opportunity  and  an  incentive  to  convince  other 
companies  of  the  actual  or  potential  value  of  its  work  and  of 
the  importance  of  supporting  it. 

To  sum  up,  it  is  my  view  that  the  society's  greatest  needs  at 
the  moment  are: 

1.  Increased  knowledge  of  the  architectural  and  decorative 
requirements  in  illumination  design. 

2.  Extension  of  educational  work  in  its  various  forms. 

3.  Successful  application  of  new  constitutional  provisions  for 
expanding  the  influence  of  the  society. 

To  make  some  progress  along  each  line  is  the  purpose  of  the 
administration.  If  material  accomplishment  is  to  result,  there 
must  be  a  continuance  of  the  active  loyal  support  of  officers, 
committees  and  members,  which  has  made  possible  the  develop- 
ments of  the  past  seven  years.  Upon  obtaining  such  support  I 
feel  that  we  may  rely  with  confidence.  Fur  the  rest,  "it  is  not  in 
mortals  to  command  success" — but  we  will  do  more,  we  will 
deserve  it.  And  paraphrasing  the  Father  of  our  Country  we 
may  say,  "Let  us  raise  a  standard  to  which  all  who  are  sin- 
cerely interested  can  repair.    The  event  is  in  the  hands  of  God." 


TRANSACTIONS  I.  F,.  S. PART  II 


1912  REPORT  OF  THE  GENERAL  SECRETARY. 


Nineteen  hundred  and  twelve,  the  seventh  year  of  the  society's 
history,  was  notable  chiefly  for  a  quiet  persistent  improvement 
in  internal  affairs  and  a  general  extension  of  the  society's  work 
along  educational  and  co-operative  lines.  The  following  brief 
review  of  the  year's  activities  records  the  salient  features. 

TECHNICAL  AFFAIRS 
Transactions. — While    there    is    some    difference    in    the    dis- 
tribution of  papers  in  the  Transactions,  yet  the  total  number 
(40)    is   so   small   as   to   make   it   impracticable   to   observe   any 
trend  in  the  character  of  the  deliberations.  A  growing  tendency  on 
the  part  of  the  Committee  on  Papers  to  refrain  from  printing 
papers  of  doubtful  permanent  value  has  reduced  the  volume  of 
the  Transactions  and  increased  the  number  of  unrecorded  lec- 
clvassificatlon  of  papers  in  first  seven  volumes  of  the 
Transactions 

Of  a  nature  to  No.  of 

No.  of  Papers         interest  particularly  Papers 

2 1         Architects 42 

Decorators 29 

Fixture  Manufacturers 28 

Ophthalmologists 27 

Manufacturers  of  Uluminants-  61 

Manufacturers       of      Lighting 

Auxiliaries 6r 

Lighting  companies 55 

Illuminating  engineers 222 

Scientists S7 


37 


Subject  of  Papers 

Light 

Physics 8 

Color 5 

General 1 

Reflection 

coefficients 7 

Uluminants 

Electric 14 

Gas •   20 

Miscellaneous 3 

Lighting  Auxiliaries  ■  ■  10 

Illumination 118 

Principles 50 

Artificial:  Interiors.   4S 

Exteriors.    16 

Natural  Outdoors  •  -     2 

Indoors  ...     2 

Units,     Standards    and 

Calculations 31 

Photometry 26 

Illuminating  Engineer- 
ing    16 

Miscellaneous 6 


Total 


26,- 


1912  REPORT  OF  THE  GENERAL  SECRETARY 

tures  or  papers  which  have  been  presented  before  section 
meetings. 

As  during  the  previous  year,  discussions  of  papers  have  often 
been  inadequate.  It  appears  advisable  to  foster  more  general 
discussions  of  papers,  this  being  the  best  safeguard  against 
inaccuracies  which  are  likely  to  creep  into  papers  in  spite  of  the 
best  precautions  of  authors,  and  papers  and  editing  committees. 
Further  emphasis  upon  advance  printing  of  papers  and  upon 
general  discussions,  whether  written  or  oral,  is  therefore  in 
order. 

The  addition  of  Volume  VII  to  the  Transactions  makes  the 
aggregate  number  of  papers  which  have  been  presented  before 
the  society  265.  These  are  distributed  as  to  character  and  scope 
substantially  as  indicated  in  the  accompanying  table. 

CONVENTION. 
One  of  the  best  evidences  of  increasing  strength  of  the  society 
is  offered  by  successive  annual  conventions,  each  one  of  which 
makes  a  more  impressive  showing  than  the  last.  The  convention 
of  1912  is  recorded  as  being  eminently  successful.  The  papers 
compared  favorably  with  those  of  previous  conventions  in 
number  and  quality;  the  maintained  attendance  at  the  sessions 
was  extraordinarily  large  in  comparison  with  the  registration, 
and  evidenced  keen  interest  in  the  proceedings.  The  spirit  mani- 
fested toward  the  society's  enterprise  was  all  that  could  be  desired. 

SECTIONS 

The  condition  of  the  established  sections  is  substantially  the 
same  as  at  the  beginning  of  the  year.  The  Philadelphia  and  New 
York  Sections  are  relatively  strong  in  interest  and  attendance  at 
meetings;  the  Chicago  Section  is  giving  evidence  of  better  main- 
tained interest  than  formerly.  The  New  England  Section  con- 
tinues in  need  of  further  development.  The  new  Pittsburgh 
Section  has  been  eminently  successful  during  its  first  year.  Its 
addition  strengthens  the  Society.  Its  potentialities  for  the  future 
seem  good. 

A  proposal  to  establish  a  new  section  at  Cleveland  has  been 
considered  during  the  year  with  adverse  conclusions.  Out  of  it 
has  grown  a  proposal  to  form  a  so-called  Middle  West  or  Lake 
Erie  Section  to  comprehend  the  large  cities,  including  Buffalo  on 


8  TRANSACTIONS  I.  E.  S. PART  II 

the  east,  Detroit  on  the  west  and  Pittsburgh  on  the  south,  meet- 
ings to  be  held  from  time  to  time  in  the  various  large  cities  of 
that  territory.  This  matter  is  still  under  consideration,  although 
it  has  made  a  favorable  appeal  in  Cleveland  and  Pittsburgh.  If 
such  a  section  is  organized  in  the  near  future,  its  effect  upon  the 
Pittsburgh  Section  will  remain  to  be  determined. 
The  statistics  for  section  meetings  are  as  follows : 

Number  Papers 

Section  of  meetings  presented 

Chicago 8  7 

New  England 6  5 

New  York 9  7 

Philadelphia  8  9 

Pittsburgh 5  7 

Totals 36  35 

COMMITTEES 

One  of  the  Society's  greatest  elements  of  strength  is  its  com- 
mittees, which  with  few  exceptions  are  gratifyingly  active  in  the 
performance  of  their  several  duties.  Approximately  75  mem- 
bers are  devoting  time  and  effort  to  the  work  of  the  general  com- 
mittees, and  others  are  serving  upon  section  committees.  The 
Society  may  well  be  proud  of  its  committee  work.  The  officers 
and  members  of  the  Council  hold  in  high  appreciation  the  ser- 
vices which  are  rendered  to  the  society  and  to  the  cause  through 
such  contributions. 

Committee  on  Papers. — Provision  of  papers  for  the  annual 
convention,  and  acceptance  or  rejection  of  all  papers  and  dis- 
cussions are  the  chief  functions  of  this  committee.  The  splendid 
convention  papers  program  attests  the  excellence  of  its  work 
during  1912. 

Committee  on  Editing  and  Publication. — The  routine  work  for 
which  this  committee  is  responsible  has  been  performed  during 
the  year  by  the  Assistant  Secretary  who  under  the  supervision  of 
this  Committee  and  of  the  Committee  on  Papers  acts  as  editor 
of  the  Transactions.  Editorial  policies  and  questions  of  print- 
ing and  distributing  the  Transactions  are  the  chief  immediate 
responsibility  of  the  committee. 

Committee  on  Section  Development. — This  committee  has 
directed  its  efforts  toward  the  improvement  and  standardization 


igi2  REPORT  OF  THE  GENERAL,  SECRETARY  9 

of  section  management  methods  and  the  extension  of  the  Society's 
influence  beyond  the  territory  in  which  the  Society  is  at  present 
represented. 

Committee  on  New  Membership. — A  new  method  of  organiza- 
tion, calculated  to  co-ordinate  the  general  and  section  new  mem- 
bership work,  was  tried  by  this  committee  during  the  year.  Due 
to  the  fact  that  section  terms  of  office  expire  in  June,  while  the 
general  society  terms  expire  in  January,  the  work  of  this  com- 
mittee has  suffered  somewhat  from  changing  personnel.  How- 
ever, a  steady,  quiet  pressure  has  been  exerted,  which  has  brought 
into  the  Society  more  than  150  men  who  are  thought  to  be 
seriously  interested  in  the  work  and  who  should  be  real  assets  for 
the  future. 

Committee  on  Reciprocal  Relations  with  Other  Societies. — 
This  committee  has  communicated  with  various  organizations 
during  the  year,  suggesting  co-operation  through  joint  meetings, 
exchange  of  representatives  on  illumination  committees,  etc. 
The  committee  has  been  very  effective  in  attaining  its  object,  and 
much  of  the  effort,  incomplete  at  this  time,  is  expected  to  bear 
fruit  during  the  coming  year.  The  lack  of  a  suitable  Society 
conspectus  has  made  the  committee's  labors  unnecessarily  arduous 
and  has  reduced  their  effectiveness  somewhat.  The  notable 
achievements  of  the  year  may  therefore  be  taken  as  an  indication 
of  larger  accomplishment  in  the  future  when  better  facilities 
shall  be  available  for  the  committee's  work. 

Committee  on  Nomenclature  and  Standards. — This,  the  sixth 
year  of  the  committee's  existence,  has  been  its  most  active  year. 
The  work  of  defining  photometrical  quantities  has  been  continued, 
and  much  of  our  photometrical  and  lighting  terminology  which 
was  in  need  of  standardization  has  been  reported  upon  during  the 
year. 

In  addition  to  this  work,  the  committee  continued  its  effort, 
undertaken  two  years  ago,  to  foster  international  co-operation  in 
standardization  of  terminology  and  units.  Its  tentative  proposal 
that  an  international  commission  be  appointed  to  deal  with  such 
matters  met  with  some  opposition  abroad  on  the  ground  that  the 
International  Photometric  Commission  (Zurich  Commission) 
provides  a  means  for  effecting  international  standardization.     A 


10  TRANSACTIONS  I.  E.  S. — PART  II 

plan  is  being  developed  for  expanding  the  scope  of  this  Commis- 
sion and  making  it  thoroughly  representative  in  every  respect. 
The  committee  has  decided  that  if  the  reorganized  commission 
fulfills  the  requirements,  no  better  auspices  could  be  obtained  for 
international  standardization  and  the  most  desirable  course 
would  be  to  await  such  reorganization  and  stand  prepared  to  co- 
operate to  the  fullest  possible  extent  in  the  international  work 
which  would  then  become  possible.  Pending  the  consummation 
of  this  plan,  all  international  endeavor  along  lines  of  standardiza- 
tion is  being  held  in  abeyance. 

Committee  on  Research. — Uncompleted  committee  organiza- 
tion and  the  lack  of  problems  demanding  immediate  attention 
have  resulted  in  a  year  of  little  activity  for  this  committee.  Con- 
siderable attention  has  been  given  to  consideration  of  the  scope, 
functions  and  personnel  of  the  committee,  and  a  preliminary  step 
was  taken  by  the  chairman  in  laying  down  the  principles  of 
research  in  a  paper  at  the  Niagara  Falls  convention. 

Committee  on  Progress. — This  committee  discharged  its  prin- 
cipal function  in  presenting  at  the  annual  convention  a  report  on 
progress  in  the  field  of  illumination. 

Committee  on  Glare  from  Reflecting  Surfaces. — Considerable 
work  in  the  way  of  study  of  its  problem  has  been  under- 
taken by  the  committee,  but  in  the  few  months  elapsing 
since  its  appointment,  it  has  been  impracticable  to  carry  much  of 
the  work  to  a  conclusion.  The  end  of  the  year  finds  the  com- 
mittee actively  engaged  in  the.  promotion  of  its  work,  making 
reappointment  of  the  present  committee  an  essential  in  order 
that  the  effectiveness  of  its  operation  may  not  be  interferred  with. 

Committee  on  Symbols. — This  committee  was  appointed  with  a 
view  to  ascertaining  whether  or  not  there  exists  a  need  which  it 
would  be  desirable  for  the  Illuminating  Engineering  Society  to 
meet  by  securing  co-operation  of  various  organizations,  in  an 
effort  to  standardize  drafting  symbols.  While  the  committee's 
final  report  is  not  yet  available,  it  appears  that  the  need  exists,  but 
there  is  considerable  doubt  if  the  Illuminating  Engineering 
Society  is  the  organization  which  should  take  the  lead  in  an 
endeavor  to  meet  it. 


IQI2  REPORT  OF  THE  GENERAL  SECRETARY  II 

Committee  on  Illumination  Primer. — The  illumination  primer, 
completed  during  the  summer,  has  been  distributed  very  generally 
among  the  lighting  fraternity,  and  those  whose  vocations  are 
such  as  to  establish  a  demand  for  knowledge  of  matters  of 
illumination.  Much  remains  to  be  accomplished  however,  in  the 
way  of  popular  dissemination.  The  primer  has  been  received 
with  encomiums  all  over  the  country  and  abroad.  It  is  expected 
that  its  influence  will  be  felt  in  the  years  to  come  both  through 
awakened  popular  interest  in  matters  of  illumination  and  through 
added  interest  in  the  society's  work. 

Committee  on  Factory  Lighting  Legislation. — The  work  of 
this  committee  represents  the  society's  first  step  toward  the  per- 
formance of  a  civic  duty.  Undoubtedly  the  New  York  State 
legislation  on  factory  lighting  will  be  the  better  for  the  influence 
which  the  committee  has  brought  to  bear  on  behalf  of  the  Society. 

Committee  on  Finance. — The  financial  condition  of  the  Society 
is  considered  and  all  payments  are  approved  at  monthly  meetings 
of  this  committee ;  also  any  financial  problems  which  the  Council 
may  refer  to  it  are  discussed  and  reported  upon.  Due  to  the 
unsatisfactory  condition  of  the  society's  finances,  this  committee 
has  been  called  upon  to  devote  more  than  usual  thought  and  time 
to  the  discharge  of  its  duties  during  the  past  year. 

Committee  on  Advertising. — It  has  been  the  desire  of  the 
Council  to  decrease  advertising  in  the  Transactions.  An  im- 
pending deficit  this  year  led  to  a  temporary  departure  from  that 
policy  and  the  committee  responded  in  a  very  gratifying  manner, 
securing  advertising  contracts  which  now  yield  a  monthly  revenue 
about  75  per  cent,  greater  than  that  obtained  at  the  beginning  of 
the  year. 

Council  Executive  Committee. — This  committee  conducted  the 
affairs  of  the  society  in  the  interims  between  council  meetings 
especially  during  the  summer  period. 

Board  of  Examiners. — This  board  is  depended  upon  to  con- 
sider and  report  on  applications  from  territories  other  than  that 
which  falls  under  section  jurisdiction,  and  to  report  to  the 
Council  upon  any  other  applications  which  may  be  referred  to  it 
for  consideration. 


12  TRANSACTIONS  I.  E.  S. — PART  II 

Committees  on  Financial  Policy  and  Revenue. — These  com- 
mittees have  considered  the  financial  problem  of  the  society  and 
have  reported  to  the  Council  upon  ways  and  means  of  meeting 
the  situation. 

Convention  Committee.- — The  notable  success  achieved  by  the 
convention  was  due  in  no  small  part  to  the  excellent  work  of  this 
committee.  Technical,  social  and  financial  details  were  handled 
in  an  expert  manner  without  calling  upon  the  society  for  financial 
assistance  other  than  the  costs  involved  in  recording  and  printing 
the  proceedings.  The  committee  financed  the  convention  and 
turned  over  to  the  society  a  small  surplus. 

Committee  on  Tellers. — Appointed  to  count  and  report  upon 
the  vote  of  the  membership  in  election  of  officers  and  on  con- 
stitutional amendments,  this  committee  discharged  its  duties  with 
promptness  and  thoroughness. 

Committee  on  Annual  Meeting. — Arrangements  for  the  annual 
meeting  as  well  as  for  the  dinner  which  preceded  it  were  entrusted 
entirely  to  this  committee  with  very  pleasing  results. 

BUSINESS  AFFAIRS. 

It  is  difficult  to  make  a  brief,  general  statement  of  the  society's 
financial  condition  because  of  accounting  difficulties  which  render 
unqualified  comparison  with  other  years  impracticable.  A  reduc- 
tion in  surplus  from  approximately  $2,900  in  January,  1912  to 
approximately  $1,450  in  January,  191 3  shows  a  decrease  of  about 
$1,450  in  surplus  as  a  result  of  the  year's  operation.  However, 
all  December  1912  expenses  have  been  included  in  the  191 2  state- 
ment whereas  about  $800  of  the  191 1  expenses  appeared  in  the 
191 2  statement.  With  this  amount  deducted  from  the  surplus 
decrease,  there  remains  an  actual  deficit  of  about  $650  for  1912. 

It  is  perhaps  not  ungratifying  to  report  this  deficit.  To  have 
reported  an  increase  in  surplus  for  the  year  would  have  been 
possible  only  as  the  result  of  a  degree  of  stagnation  of  society 
affairs.  Some  of  the  new  activities  which  commended  themselves 
to  the  Council  were  undertaken  in  the  face  of  an  assured  deficit 
because  they  were  considered-  well  worth  while  notwithstanding 
the  unfavorable  financial  aspect.  The  inadequacy  of  members' 
dues  to  meet  the  ordinary  expenses  has  been  recognized  by  the 
Council  and  reported  to  the  membership  during  the  past  few 


10,12  REPORT  OF  THE  GENERAL  SECRETARY  13 

years.  The  deliberate  sense  of  the  Council  has  been  that  since 
the  society's  enterprises  were  worthy  of  support,  they  should  be 
carried  out  and  increased  support  should  be  obtained,  instead  of 
curtailing  enterprise  to  the  point  where  the  support  already  avail- 
able would  prove  adequate. 

The  year's  expenses  include  approximately  $1,000  involved  in 
the  production  and  dissemination  of  the  illumination  primer.  If 
it  shall  be  found  feasible  to  arrange  for  the  extensive  dissemina- 
tion of  the  primer  on  terms  which  will  make  it  possible  for  the 
society  to  reimburse  itself  for  this  $1,000  expenditure,  the  year's 
operations  may  be  recorded  as  showing  a  profit  rather  than  a  loss. 

MEMBERSHIP. 
There  has  been  some  reduction  in  the  number  of  members  dur- 
ing the  year,  the  defections  being  larger  than  the  additions  to  the 
roll.     The  statistics  are  as  follows : 

Members  January  1 ,  1 9 1 2 1,418 

Additions  during  year 206 

Defections  during  year 289 

Membership  December  31,  1912 1.335 

Seven  members  have  been  removed  from  us  by  death.  Their 
names  are : 

George,  Thomas  L. 

General  Superintendent,  The  United  Gas  Improvement  Co.,  Broad 

and  Arch  Streets,  Philadelphia,  Pa. 
Manning,  Wm.  J. 

District  Manager,  Philadelphia  Electric  Co.,  Gray's  Ferry  Road  and 

Carpenter  Street,  Philadelphia,   Pa. 
Mayer,  Frederick  J. 

General  Manager,  Didier  March  Co.,  30  Church  Street,  New  York, 

N.  Y. 
McGlensey,  J.  F. 

Illuminating  Engineer,  Union  Electric  Light  &  Power  Co..  12th  and 

Locust  Streets.  St.  Louis,  Mo. 
Morgan,  A.  J. 

Secretary,   National   X-Ray   Reflector  Co..   235   W.  Jackson   Boule- 
vard, Chicago,  111. 
Riblet,  A. 

Superintendent,    14th    Street    Station,    Consolidated    Gas    Co.,    14th 

Street  and  Avenue  C,  New  York,  N.  Y. 
Spillman,  A.  J. 

Chief    of    Lamp    Dept.,    Philadelphia    Electric    Co.,    1000    Chestnut 

Street,  Philadelphia,  Pa. 


14 


TRANSACTIONS  I.  £.  S. — PART  II 


j      co 

M 

CO 

>o 

~t 

0 

t^ 

^5 

t^  VO 

^o 

ON 

m  on  co    >* 

w 

■*  tt 

LO 

>-       Tl- 

tN 

u  +* 

Co 

e 

O 

M 

■>* 

lO 

M 

O 

o    o 

cr. 

-r 

If}    tO    N    CO 

■* 

0     ■* 

CN     o 

lO 

CO      O 

o    M 

tN 

CO 

o 

CM 

rT3 

tN 
tN 

co 

ON  CO 

tN 

tN 

ON 

■*  »D     N    N 
r^  CO     rD    i-i 

to 

ir>   on 

O 

tN 

CO    M3 
tN 

o 

m  Tt-  vo    w    I    tj-    I      I    vo    I    «'  I    «o 


m     rO    O     tN     tN 


VO     O 


to    tN     r-»    M     On 


<S  td 
21 


w     rO    O     tN 


rOCOlOrOfNt^lOl       ■*!       H       I       r>. 


O 
CO 

o 

M 
« 

w 
w 

t-H 

o 

w 

o 

fcH 

«) 

M 

p 


M     to    to    m 


^ONtN       I       t>.Mt^NMl-ICM«        I       Is* 


1/JV3      IO    f)    H      pi      N      H 


tfl 


-23     rt 


3  '5  1 


3    £ 


J3  u 

t  s 

<2    S  So 

«)   io  n 

53   2   a  -5 

&^  *  s 

O     "J    CvO  a 

M  -9   a  "3 

J3  "£  "^  a 

u  ft  «  ■-< 


bo 

.9 

be  a 
3  3 

O     fcuO 

£  '3 


bjO  ? 

a  £ 


.0 

CJ 

3 

hH 

U 

I) 

Ct 

K 

O 

Tj 

a   cs 


t-I    ca    tri 


y  M 


5  - 
'5c 


a   a 


hi  e-5 


W  .2 

d  j 

co   a 
.   a 

M 

'C    to 

c   a  _ 
,  <3   o  y. 

•S  a  «g 

ft  cd  s 

•a  "  « 

.—  ^3  « 

a   <->  v, 

a  . 


HH^P 


0) 

O 
en     s- 

£     ft 


1912  REPORT  OF  THE  GENERAL  SECRETARY  15 

The  membership  record  is  involved  by  the  fact  that  bills  for 
1913  dues  were  issued  early  in  December  and  resulted  in  a  num- 
ber of  resignations  which  otherwise  would  not  have  been  received 
until  January.  Resignations  are  always  to  be  expected  when 
bills  for  dues  are  issued.  In  this  1912  record,  the  defections  in- 
clude such  resignations  for  the  beginning  and  the  end  of  the 
year,  while  in  other  years  they  include  such  resignations  only  for 
the  beginning  of  the  year.  Corrected  for  this  irregularity,  the 
membership  is  found  to  have  about  held  its  own. 

An  approximate  classification  of  the  members  by  vocation 
appears  in  the  preceding  table.  Such  a  classification  is  difficult 
and  liable  to  unavoidable  error  so  that  the  distribution  should  be 
considered  no  more  than  indicative. 

GENERAL  OFFICE. 
The  last  annual  report  noted  the  inadequacy  of  office  facilities 
in  the  United  Engineering  Societies  Building,  New  York.  Dur- 
ing the  past  year  an  adjoining  room  has  been  secured  which  is 
now  used  as  a  business  office,  while  the  original  room  serves  as 
the  private  office  of  the  Assistant  Secretary  for  editorial  and 
other  work,  and  for  the  meeting  room  of  the  Council,  various 
committees,  and  the  New  York  Section  Board  of  Managers. 

FOREIGN  RELATIONS. 
Abroad  the  year  has  been  signalized  by  the  announcement  of 
the  formation  of  a  German  Illuminating  Engineering  Society 
(Deutsche  Beleuchtungstechnische  Gesellschaft).  Anticipating 
that  the  organization  of  this  Society  will  be  perfected  during  the 
coming  year,  it  may  be  expected  that  the  three  societies  located  in 
respectively  the  United  States,  England  and  Germany,  will  find 
many  points  of  common  interest  and  much  that  is  mutually 
beneficial  in  their  activities.  With  the  probability  that  the 
reorganized  International  Photometric  Commission  will  meet  the 
requirements  of  international  standardization,  there  remain  for 
development  methods  of  co-operation  in  other  matters  among 
the  various  societies.  The  Illuminating  Engineering  Society 
through  its  Council  stands  ready  to  co-operate  to  the  fullest 
extent  with  the  English  and  German  societies  through  their 
governing  boards. 
3 


l6  TRANSACTIONS  I.  E.  S. — PART  II 

GENERAL. 

The  year's  experience  has  indicated  some  lack  in  internal 
organization.  Primarily  this  will  be  met  if  there  can  be 
established  closer  co-operation  among  the  sections  and  between 
the  sections  and  the  Council.  In  this  latter  work  the  respective 
Vice-Presidents  can  be  very  influential.  Among  the  committees 
there  has  been  evidenced  a  need  for  better  organization.  Most 
committees  need  to  hold  at  least  occasional  meetings,  if  the 
committee's  organization  is  to  be  perfected  and  its  plans  definitely 
formulated.  With  these  two  general  exceptions  the  organization 
of  the  society  appears  to  be  satisfactory. 

The  opportunities  for  effective  service  loom  large  for  the 
future.  Provisions  for  making  available  additional  funds  with 
which  to  carry  out  the  work  of  the  society  should  remove  the 
only  serious  handicap  under  which  we  have  labored  in  the  recent 
past.  The  proven  loyalty  of  large  numbers  of  members  gives 
assurance  of  means  of  accomplishment.  All  things  considered, 
there  is  every  reason  to  look  for  further  growth  and  extension  of 
influence  in  the  near  future. 

Preston   S.   Millar, 

General  Secretary. 


shaleing:    department  store  lighting  17 

DEPARTMENT  STORE  LIGHTING.* 


BY   H.    W.   SHAELING. 


Synopsis: — This  paper,  while  it  is  for  the  most  part  confined  to  the 
re-designing  of  a  particular  lighting  installation,  outlines  the  problem 
usually  encountered  in  the  lighting  of  large  department  stores.  Lighting 
systems,  the  relative  advantages  of  direct  and  indirect  lighting,  uniformity 
of  illumination,  color  of  light,  avoidance  of  objectionable  shadows,  main- 
tenance, etc.,  questions  which  demand  consideration  in  problems  of  this 
sort  are  discussed  briefly.  Generally,  the  re-designing  in  this  case  con- 
sisted of  the  substitution  of  tungsten  lamp  units  for  an  existing  installa- 
tion of  electric  arc  and  carbon  incandescent  lamps.  Both  the  old  and 
the  new  installations  are  described  in  detail.  Photometric  data,  results 
of  illumination  tests  made  on  several  floors,  diagrams  of  the  test  stations 
and  photographs  of  the  two  installations  are  also  included.  The  installa- 
tion changes  described  have  afforded  a  lighting  system  which  is  not  only 
more  artistic  and  effective,  but  one  that  has  reduced  the  operating  costs 
of  the  old  system  more  than  fifty  per  cent. 

The  choice  between  different  systems  and  different  lighting 
units  is  generally  made  on  the  basis  of,  (1)  relative  efficiency; 
(2)  relative  attractiveness  in  appearance.  These  are  the  funda- 
mental considerations,  although  there  are  others  which  are  also 
of  vital  importance. 

LIGHTING  SYSTEMS. 

All  lighting  systems  can  in  general  be  classed  under  one  of 
three  classifications,  viz.,  direct,  indirect,  and  semi-indirect. 

In  practically  all  lighting  systems  some  portion  of  the  illumina- 
tion is  received  indirectly.  In  other  words  a  portion  of  the 
illumination  is  obtained  by  light  reflected  from  the  ceiling  or 
walls,  or  both,  before  reaching  the  plane  of  utilization.  In  direct 
lighting  when  efficiency  is  important,  it  is  the  aim  in  general  to 
make  the  indirect  portion  of  the  illumination  small,  allowing  only 
enough  light  to  reach  the  ceiling  and  walls  to  illuminate  them 
to  a  low  intensity,  thereby  preventing  a  gloomy  appearance. 

Indirect  illumination  is  produced  by  the  light  coming  from  a 
very  large  area — the  ceiling  and  upper  portion  of  walls.     This 

*  A  paper  read  before  a  meeting  of  the  Pittsburgh  section  of  the  Illuminating 
Engineering  Society,  January  24,  1913. 


1 8  TRANSACTIONS  I.  E.  S. — PART  II 

gives  what  is  known  as  diffuseness  of  illumination.  In  such  a 
system  no  direct  light  is  received  on  the  plane  of  utilization,  the 
light  source  being  concealed  in  an  opaque  unit. 

What  has  been  said  in  regard  to  indirect  lighting  applies  equally 
well  to  semi-indirect  lighting  with  this  exception,  viz.,  that  the 
light  source  is  mounted  in  a  translucent  rather  than  an  opaque 
unit  so  that  some  of  the  illumination  is  received  directly.  When 
properly  designed,  a  semi-indirect  system  possesses  all  the  illu- 
mination advantages  of  totally  indirect  lighting.  In  addition,  it 
is  more  attractive  in  appearance  and  avoids  the  unpleasant  effect 
of  a  brilliant  ceiling  with  no  visible  source  of  light.  The  great 
danger  in  the  use  of  semi-indirect  lighting  is  that  the  translucent 
units  used  will  transmit  too  much  light.  When  too  much  light  is 
transmitted,  the  efficiency  is  very  rarely  any  greater  than  with 
totally  indirect  lighting,  and  the  illumination  advantages  of  the 
indirect  illumination  are  greatly  reduced. 

Experience  has  shown  that  the  best  degree  of  transmission  of 
light  with  semi-indirect  lighting  units  is  possible  when  the  bril- 
liancy of  the  light  unit  is  approximately  the  same  as  the  brilliancy 
of  the  ceiling. 

INDIRECT  COMPARED  WITH  DIRECT  LIGHTING. 

Obtaining  a  large  portion  of  the  illumination  indirectly  has  the 
following  disadvantages  as  compared  with  direct  lighting. 

(i)  Lower  efficiency;  to  produce  a  given  illumination  requires 
about  twice  as  much  light  with  indirect  lighting  as  with  efficient 
direct  lighting. 

(2)  More  rapid  deterioration  due  to  the  collection  of  dirt. 

(3)  A  lower  degree  of  perspective,  since  sharp  shadows  are 
largely  eliminated. 

(4)  An  unduly  bright  ceiling  which  often  gives  an  unpleasant 
psychological  effect,  especially  when  the  opaque  unit  of  the  in- 
direct lighting  forms  a  contrast  with  the  brightly  lighted  ceiling. 

For  the  reasons  cited  above,  and  since  in  the  average  depart- 
ment store  a  comparatively  large  area  must  be  illuminated,  it 
would  seem  that  in  most  cases  the  general  illumination  can  be 
obtained  more  economically,  efficiently,  and  with  units  which  will 
be  sufficiently  attractive  in  appearance,  by  a  system  of  direct 
illumination. 


shalung:    department  store  lighting  19 

Moreover,  it  is  generally  considered  by  most  authorities  that 
the  advertising  value  or  attractive  power  of  a  direct  is  far  greater 
than  that  of  an  indirect  system.  It  is  not  meant  by  this,  however, 
that  exposed  light  sources  should  be  used  for  this  purpose,  but 
either  totally  enclosing  units  or  reflectors  which  practically  con- 
ceal the  light  sources  in  the  case  of  incandescent  electric  lights. 
It  is  not  the  purpose  of  a  department  store  to  install  lighting 
fixtures  for  display,  except  in  those  spaces  as  may  be  devoted 
to  the  sale  of  such  goods.  The  light  units  should  not  attract 
attention  from  the  goods  displayed;  the  object  should  be  to 
provide  proper  illumination  for  the  display  of  goods  at  a  reason- 
able expense  to  the  owner;  the  fixtures,  of  course,  should  be 
sufficiently  artistic  in  appearance  and  of  a  character  which  will 
be  in  harmony  with  the  architectural  surroundings. 

IMPORTANT  CONSIDERATIONS  INVOLVED. 

Color  Value. — The  light  units  employed  should  give  light  which 
in  color  approaches  as  near  as  possible  that  of  natural  daylight, 
so  as  not  to  distort  the  colors  of  the  goods  displayed.  But  there 
are  certain  classes  of  goods  which  will  be  used  almost  entirely 
under  artificial  light  such  as  is  found  in  the  home,  theatre,  and 
similar  places;  the  light  under  which  these  goods  are  sold  should 
approximate  that  under  which  goods  will  be  used,  which  is  in 
general  the  incandescent  electric  lamp. 

Due  to  its  high  efficiency,  the  possibility  of  a  more  efficient 
utilization  of  its  light  by  the  use  of  properly  designed  reflectors, 
ease  of  maintenance,  and  the  range  in  sizes  available,  the  tungsten 
filament  lamp  has  gradually  displaced  the  arc  lamp  for  depart- 
ment store  lighting.  While  tungsten  lamps  do  not  give  the  same 
color  values  as  are  given  by  daylight,  the  approximation  is  close 
enough  for  most  practical  purposes.  For  particular  cases,  special 
arc  and  incandescent  units,  which  give  a  closer  approximation 
to  daylight  values,  have  been  developed  and  are  continually  being 
improved.  At  the  present  time,  however,  they  are  being  used  in 
specially  prepared  spaces  and  not  for  general  purposes. 

Avoidance  of  High  Intrinsic  Brilliancy. — The  avoidance  of 
glare  from  exposed  brilliant  sources  is  essential.  If  enclosing 
opal  or  prismatic  glassware  be  used,  this  effect  is  reduced  prac- 
tically to  a  minimum.     If  prismatic  or  opal  reflectors  are  used 


20  TRANSACTIONS  I.  E.  S. — PART  II 

they  should  be  of  a  deep  bowl  shape  so  as  to  completely  screen 
the  lamp  filament  from  the  eye  in  all  its  normal  positions.  It 
is  advisable  also  that  the  lower  portion  of  the  bowl  of  the  lamp 
be  etched  or  frosted. 

Uniformity  of  Illumination. — The  light  units  should  afford 
such  distribution  of  light  as  will  produce  a  reasonable  degree  of 
uniformity  of  illumination ;  glassware  which  will  effect  the  most 
desirable  distribution  of  the  light  from  the  particular  lamps,  or 
for  any  particular  arrangement  of  outlets  or  class  of  lighting 
service  should  be  selected  with  care. 

Avoidance  of  Objectionable  Shadows. — It  is  highly  important 
to  avoid  objectionable  shadows  cast  by  a  customer  in  standing 
before  a  counter.  For  this  reason  careful  attention  should  be 
paid  to  the  arrangement  of  light  units  and  their  mounting  heights. 

Efficiency. — Efficiency  is  here  used  in  the  sense  of  illuminating 
efficiency.  The  reflector  type  of  glassware  is  generally  consid- 
ered more  efficient  than  the  enclosing  type,  and  is  to  be  recom- 
mended wherever  the  consideration  of  efficiency  is  of  primary 
importance.  Enclosing  glassware  is  to  be  recommended  where 
maximum  attractiveness  in  appearance  is  more  important  than 
efficiency. 

Cleaning  and  Maintenance. — Any  efficient  lighting  system 
requires  careful  attention  to  cleaning  and  maintenance.  Under 
conditions  such  as  exist  in  a  large  department  store  as  well  as 
in  any  first  class  lighting  installation,  the  reflectors  or  enclosing 
globes  must  be  periodically  inspected  and  cleaned,  or  the  appear- 
ance and  the  illuminating  efficiency  of  the  installation  will  be  seri- 
ously impaired.  Lighting  installations  require  regular — though 
perhaps  less  frequent — cleaning,  just  as  do  the  store  windows. 
It  will  be  shown  later  how  in  one  instance  this  detail  has  been 
taken  care  of  in  a  systematic  manner  and  at  a  very  slight  cost. 

THE  McCREERY  &  COMPANY  (PITTSBURGH)  STORE 

In  the  summer  of  1912  McCreery  &  Company  decided  to  re- 
design the  lighting  equipment  of  their  store,  which  is  located  at 
Wood  Street  and  Sixth  Avenue,  Pittsburgh,  Pa.  At  that  time 
their  lighting  equipment  consisted,  for  the  main  part,  of  enclosed 
carbon  arc  lamps  with  a  considerable  number  of  carbon  filament 


\* 


Fig.  i. — Former  lighting  equipment  of  first  floor. 


Fig.  2.- -Night  view  showing  lighting  units  on  first  floor. 


Fig.  3.— Third  floor  under  old  lighting  system. 


pig,  4.— lighting  unit  used  on  the  second,  third,  fourth,  fifth,  sixth,  seventh  and  twelfth  floors. 


shalling:    department  store  lighting 


21 


and  a  few  tungsten  incandescent  lamps.  Their  purpose  in 
re-designing  their  lighting  equipment  was,  first,  to  improve  their 
lighting  by  providing  a  higher  intensity  and  more  uniform  illu- 
mination; second,  to  reduce  their  lighting  expenses,  and,  third, 
to  provide  a  modern  and  attractive  installation. 

The  building  is  14  stories  high,  including  the  basement  and  the 
attic.  The  ground  floor  covers  a  space  100  ft.  x  216  ft.  (30.48  m. 
x  75.84  m.)  and  each  of  the  remaining  floors  90  ft.  x  216  ft. 
(27.43  m.  x  75.84  m.).  The  inside  area  of  the  first  floor  is 
approximately   20,370   square    feet    (1892.44   sq.    m.),   and   the 


Fig.  5.— Photometric  curve  of  the  lighting  unit  (with  a  250-watt  clear 
tungsten  lamp)  shown  in  fig.  4. 


remaining  floors  17,850  square  feet  (1658.32  sq.  m.).  The  out- 
side of  the  building  is  finished  in  white  terra-cotta;  the  interior, 
including  the  fixtures,  is  of  mahogany. 

A  careful  investigation  was  made  of  practically  all  types  of 
lamps  and  reflecting  devices  with  the  view  of  determining  their 
relative  efficiencies  and  applicability  for  the  various  spaces  to  be 
lighted.  As  a  result  of  the  investigation,  tungsten  lamps  with 
various  accessories,  which  are  described  further  on  in  the  paper, 
were  decided  upon  for  the  different  departments. 

All  changes  were  made  with  practically  no  changes  in  wiring; 


22  TRANSACTIONS  I.  E.  S. — PART  II 

all  the  necessary  work  such  as  removing  the  old  and  hanging 
new  fixtures  was  taken  care  of  by  the  electrical  department  of 
the  store,  with  no  outside  assistance  whatever.  One  feature 
worthy  of  particular  attention  and  which  might  well  be  followed 
by  others  in  making  such  changes,  is  the  fact  that  all  the  changes 
on  each  floor  were  made  in  a  single  night,  so  that  when  the  sales 
force  arrived  the  next  morning  they  found  a  complete  new  light- 
ing equipment  with  no  trace  of  the  dirt  incident  to  the  change. 
In  this  way  business  was  not  interrupted. 

EQUIPMENT  OF  VARIOUS  FLOORS  OF  THE  BUILDING. 

Basement  (Shoes,  trunks  and  bags). — The  lighting  equipment 
consisted  of  150  and  250- watt  bowl-frosted  tungsten  lamps  and 
prismatic  satin  finish  reflectors  of  the  extensive  type  suspended 
from  ceiling  by  single  pendant  drops  15  in.  (0.38  m.)  in  length. 
Ample  and  satisfactory  illumination  was  obtained  in  this  space 
and  no  changes  were  made. 

Basement  (Delivery  department). — The  equipment  on  this 
floor  consisted  of  clusters  of  carbon  filament  lamps  which  were 
replaced  by  18  single  units  consisting  of  a  60- watt  clear  tungsten 
lamp  and  porcelain  enameled  shallow  dome  type  steel  reflector. 

First  Floor  (Men's  furnishings,  jewelry,  leather  goods,  station- 
ery, dress  trimmings,  toilet  articles,  etc.). — The  lighting  equip- 
ment of  this  floor  consisted  of  34  2-light  fixtures  and  16  i-light 
fixtures  supporting  5-ampere  enclosed  arc  lamps,  and  5  4-light 
decorative  brackets  on  pillars  at  the  rear  of  the  store  equipped 
with  all-frosted  carbon  lamps. 

In  view  of  the  fact  that  ornamental  and  expensive  fixtures 
were  already  installed,  an  equipment  of  lighting  accessories  in 
harmony  with  them  was  selected.  A  reflector-ball  type  of  unit  of 
special  design  was  decided  upon ;  it  consists  of  an  upper  prismatic 
reflector  resting  on  a  stalactite  shaped  blown  globe,  the  entire 
unit  being  satin  finished.  A  noteworthy  feature  of  this  globe 
is  the  fact  that  distributions  approximating  the  \Vell-known 
extensive,  intensive,  and  focusing  distributions  may  be  obtained 
by  varying  the  position  of  lamp  within  the  globe.    A  position  was 


»* 


Fig.  6.— Night  view  of  lighting  units  on  third  floor. 


Fig.  7.— Scheme  for  lighting  rug-racks  on  the  fifth  floor. 


Fig.  8.  — Eighth  floor  with  old  lighting  system. 


Fig.  9.— Night  view  of  eighth  floor  showing  new  lightim 


shalung:    department  store  lighting  23 

determined  upon  which  would  give  practically  an  intensive  dis- 
tribution. 

Each  arc  lamp  was  replaced  by  a  250-watt  clear  tungsten  lamp. 
The  decorative  units  on  the  pillars  were  replaced  by  25-watt 
bowl-frosted  tungsten  lamps  and  decorative  shades  having  a 
satin  finished  plate  over  the  bottom,  with  the  exception  of  the 
space  occupied  by  lamp  bulb. 

Second  Floor  (Bedding  and  yard  goods). — The  equipment  on 
this  floor  formerly  consisted  of  53  5-ampere  direct  current 
enclosed  arc  lamps  with  clear  inner  and  opal  outer  globes.  A 
fairly  high  intensity  of  illumination  is  required  on  this  floor  for 
matching  dress  goods  and  similar  purposes.  Each  arc  lamp  was 
replaced  by  a  400- watt  clear  tungsten  lamp  placed  within  a  14  in. 
(35.56  cm.)  2-piece  diffusing  glass  bowl.  This  unit1  (see  Fig.  4) 
consists  of  a  clear  stiletto  prism  reflector  mounted  over  a  satin 
finished  shallow  bowl,  smooth  on  the  outside  and  having  stiletto 
prisms  on  the  interior  surface.  These  units  are  mounted  in  very 
attractive  fixtures  having  a  verde  antique  finish.  The  overall 
length  of  the  fixture  is  42  in.  (1.07  m.)  so  that  the  bottom  of 
unit  is  approximately  10  ft.  6  in.  (3.20  m.)  above  the  floor. 

Third  Floor  (Ladies'  suits,  waists,  and  furs). — Fifty-one  arc 
lamps  on  this  floor  were  each  replaced  by  a  250-watt  clear 
tungsten  lamp  unit  and  fixture  similar  to  those  used  on  the 
second  floor  (Fig.  4).  A  number  of  side  bracket  units  equipped 
with  60-watt  clear  carbon  lamps  were  replaced  by  25-watt 
tungsten  lamps. 

Fourth  Floor  (Millinery,  ladies'  and  infants'  wear). — In  the 
millinery  department  of  this  floor  12  arc  lamps  were  each  replaced 
by  a  semi-indirect  unit  equipped  with  a  150-watt  clear  tungsten 
lamp.  These  units  are  of  a  design  to  -correspond  to  the  archi- 
tecture of  the  room  which  is  of  French  design  of  the  period 
of  Louis  XIV.  A  soft  diffused  light  is  obtained,  which  is  supple- 
mented to  a  considerable  extent  by  the  illumination  received  from 
the  units  used  for  lighting  the  millinery  show  cases.    Thirty-three 

1  A  "  Holophane-Realite  "  (shown  in  Fig.  4.). 


24 


TRANSACTIONS  I.  E.  S. — PART  II 


arc  lamps  on  the  main  portion  of  this  floor  were  each  replaced  by 
a  unit  such  as  used  on  the  third  floor  (Fig.  4). 

Fifth  Floor  (China,  bric-a-brac  and  rugs). — Fifty  arc  lamps 
were  formerly  used  on  this  floor.  Twenty-five  were  replaced  by 
units  like  those  used  on  the  second  floor  (see  Fig.  4)  equipped 
with  150-watt  clear  tungsten  lamps,  and  twenty-five  by  similar 
units  equipped  with  250-watt  clear  tungsten  lamps. 

A  special  form  of  lighting  was  installed  for  the  rug  rack  (see 
Fig.  7).    Twenty-five  outlets  were  provided  on  the  circumference 


Vr^iWf 

iillP^ 

>kw^\vw\\\\V>cW 

iMttmmm 

PMlilli 

wM0m 

ill  1 1 1   1  till 

1111 

|| 

iUwuT\VQ-«T 

i3^^^^^% 

ic^ll?i^$&; 

llMm? 

-^s^N^t: 

^^^ 

Fig.  10. — Photometric  curve  of  eight-inch  light-density  opal*  reflector  with 
60-watt  bowl-frosted  tungsten  lamp. 


of  a  semi-circle,  the  radius  of  which  was  approximately  2  ft. 
(0.61  m.)  greater  than  the  maximum  swing  of  the  rack  arms,  the 
center  of  the  rack  being  considered  the  center  of  the  semi-circle. 
Outlets  were  located  on  2  ft.  6  in.  (0.77  m.)  centers.  Wiring  was 
run  in  metal  moulding  and  small  porcelain  receptacles  having 
a  metallic  bead  for  use  with  shade  holder  were  used.  Each  outlet 
was  equipped  with  one  40-watt  clear  tungsten  lamp  and  a  30-deg. 
angle  steel  reflector  having  an  aluminumized  interior  surface. 
The  units  were  mounted  on  the  ceiling  in  a  pendant  position.  In 
order  that  the  light  units  should  not  present  too  strong  a  contrast 

*  "Veluria"  reflector  of  the  Holoplane  Works  of  the  General  Electric  Co. 


shalling:    department  store  lighting 


25 


with  the  color  of  the  ceiling,  the  exterior  surfaces  of  reflectors 
were  also  finished  with  aluminum. 

Sixth  Floor  (Mission  furniture,  wall  paper,  draperies,  art 
goods,  and  ladies'  parlor). — Thirty-five  arc  lamps  used  for  the 
lighting  of  the  main  portion  of  the  floor  were  each  replaced  by 
units  similar  to  those  used  on  the  third  floor.  The  ladies'  parlor 
was  equipped  with  four  5-light  fixtures  using  carbon  lamps. 
The  fixtures  were  retained  and  15  and  40- watt  clear  tungsten 
lamps  in  6  in.  (15.24  cm.)  and  7  in.  (17.78  cm.)  prismatic  balls 
were  substituted  for  carbon  lamps.    In  the  art  rooms  which  are 


Fig.  11.— Photometric  curve  of  60- watt  clear  tungsten  lamp  and  concentrating 
type  reflector  used  in  show  windows. 


decorated  in  an  oriental  manner,  five  light  fixtures  which  were  in 
use  were  retained;  25-watt  tungsten  lamps  were  substituted  for 
60-watt  carbon  lamps  and  placed  within  decorative  shades  similar 
to  those  used  in  the  brackets  on  the  first  floor. 

Seventh  Floor  (Men's  and  boys'  clothing  and  general  offices). — 
The  general  illumination  of  this  floor  was  taken  care  of  by  45 
arc  lamps,  36  of  which  were  replaced  by  400-watt  and  9  by 
250-watt  clear  tungsten  lamps  in  special  units  like  that  shown  in 
Fig.  4.  As  on  the  second  floor,  a  high  intensity  of  illumination 
is  required,  on  account  of  the  nature  of  the  goods  displayed.  A 
feature  which  was  introduced  on  these  two  floors  to  obtain  a 


26 


TRANSACTIONS  I.  E.  S. PART  II 


whiter  light  and  allow  greater  facility  in  the  matching  of  colors, 
was  the  use  of  lamps  rated  three  volts  below  the  circuit  voltage. 
It  is  interesting  to  note  in  this  connection  the  excellent  results 
obtained,  also  the  fact  that  even  though  the  lamps  are  burned 
above  their  rated  voltage,  a  life  of  800  hours  is  obtained. 

Offices — 21  ceiling  outlets  equipped  with  100-watt  bowl-frosted 
tungsten  lamps  with  flat  opal  shades  and  24  2-light  desk  standards 
were  replaced  by  21  ceiling  outlets  equipped  with  150-watt  bowl- 
frosted  tungsten  lamps  and  intensive  type  reflectors. 


y4  ~  7y/>/ca/  secf/o/t  of '<sAo*y  wxitfows  ^Aotv/'ny  /ocof/o/j  of  /.tfAf  i/w'fa. 

•      /nef/cofes  /ocof/o/r  of  Co/t/m/ra. 

©     Ihcftcofso  fecaf/o/?  of  2-*}f>f  i/nite 
X~  Intf/ccfes  Soy  /njvtiicA:  ///vm/Mf/aft^ff'ewf//?^  ttrere  ro/rer>. 

Fig.  12.— Plan  of  first  floor  showing  location  of  furniture, 
pillars,  light  units,  etc. 


Eighth  Floor  (Furniture  and  victrola  department). — This 
floor  was  formerly  lighted  by  85  6-light  cluster  units  consisting 
of  60-watt  all-frosted  carbon  filament  lamps  mounted  under  a 
large  porcelain  shade.  In  the  main  portion  of  this  floor  80  units 
were  replaced  by  a  single  60-watt  bowl-frosted  tungsten  lamp 
and  an  8  in.  (20.32  cm.)  opal  reflector  fastened  in  the  present 
fixtures. 

The  victrola  department  is  now  illuminated  by  five  150-watt 
clear  tungsten  lamps  in  semi-indirect  lighting  units. 

Ninth  Floor  (Dining  room  and  kitchen). — In  the  dining  room 
71  6-lamp  cluster  fixtures  had  been  installed;  each  lamp  was  a 
30-watt  clear  carbon  filament  lamp.    The  fixtures  were  retained; 


shalling:    department  store  lighting 


27 


15-watt  all-frosted  tungsten  lamps  were  used  to  replace  30-watt 
carbon  lamps.  Numerous  minor  changes  were  also  made  on  this 
floor;  60-watt  carbon  lamps  were  replaced  by  25-watt  tungsten 
lamps. 

In  the  kitchen  11  arc  lamps  were  replaced  by  250-watt  clean 
tungsten  lamps. 


S- 


e  t© 


I 


-^e- 


^ 


-s'f 


®  ©  © 


370     S.SO  SO  ■ 

n&r*  ©  * 

stz  1  lsao  *t; 

i 

*2t\L*Jf  M 
.Vt 

£08      ***  40S 


I 


©      © 


twin 


//x//'co/es  /oca f '/on  of  7es/ S/of/ons.        Ce///nff  /re /$/}/,   cfO'-O 
9  /mf/ca/ej  /oca//on  0/  Co/i/mns.  f/e/pn/ of '/amps  cboye  f/oor,  \tff-6 

®  frtd/co/es  /oca/zon  of  //$///  C/n//s. //e/$h/  of  /amps  adore  7ej/  f/one,  9'-/Q 


Fig.  13.— Section  plan  of  first  floor  showing  location  of  illumination  test  stations. 


Tenth  Floor  (Furniture,  employees'  recreation  and  lunch 
room). — Thirty-two  arc  lamps  were  replaced  by  150- watt  clear 
tungsten  lamps  equipped  with  extensive  type  prismatic  reflectors. 
Seventy-six  60-watt  clear  carbon  lamps  were  each  replaced  by  a 
25-watt  tungsten  lamp. 

Eleventh  Floor  (Buyers'  offices,  receiving  and  stock  rooms). — 
The  equipment  formerly  consisted  of  19  arc  lamps  and  13  4-light 


28 


TRANSACTIONS  I.  E.  S. — PART  II 


clusters  of  6o-watt  clear  carbon  filament  lamps.  Thirteen  of  the 
arc  lamps  were  replaced  by  250-watt  and  6  by  150-watt  tungsten 
lamps.  Each  4-light  cluster  was  replaced  by  a  single  60-watt 
tungsten  lamp  and  extensive  type  prismatic  reflector. 

Twelfth  Floor  (Fitting  rooms  and  wash-rooms). — Forty  arc 
lamps  were  formerly  used.  Ten  were  replaced  by  250-watt 
tungsten  lamps  in  the  special  unit  shown  in  Fig.  4  and  30  by 
150-watt  tungsten  lamps  and  opal  shades. 

Thirteenth  Floor  (Attic). — Six  arc  lamps  and  100  60-watt 
carbon  drop  lights  were  formerly  used;  4  of  the  arc  lamps  were 
replaced  by  150-watt  tungsten  lamps,  and  2  by  100-watt  tungsten 


#     /nd/cates   /ocof/on  of  Co/vmns. 

©      Jnd/'cates  /ocof/on  of  £./gt>f  Umfs. 

X  "  Jhdicofes-bqy  in  tv/i/c/r  Ifft/minof/on  ffeod/nga  ttvse  fo/zen. 

Fig.  14.— Plan  typical  of  the  general  arrangement  of  furniture,  light  units,  etc., 
on  the  upper  floors,  with  exception  of  eighth. 

lamps.  Each  60-watt  carbon  lamp  was  replaced  by  a  25-watt 
tungsten  lamp. 

Show  Window  Illumination. — An  average  section  of  these  win- 
dows is  23  ft.  6  in.  (7.16  m.)  long  and  8  ft.  (5.49  m.)  deep,  the 
ceiling  is  19  ft.  (5.79  m.)  above  the  bottom  of  the  show  case. 
The  plate  glass  extends  to  a  height  of  12  ft.  (3.66  m.).  Prism 
glass  being  used  at  the  top.  The  sides  and  back  of  the  windows 
up  to  a  height  of  six  feet  are  covered  by  a  dark  green  velvet 
curtain  or  drapery ;  clear  glass  plates  are  used  above  this  to  let  as 
much  daylight  into  the  store  as  possible. 

The  former  equipment  for  windows  consisted  of  a  total  of 
326  60-watt  clear  carbon  lamps  mounted  in  a  mirrored  trough 
placed  back  of  the  transom  bar.    The  present  equipment  consists 


shalling:     department  store  LIGHTING 


29 


of  a  total  of  195  60- watt  clear  tungsten  lamps;  15  are  used  in  each 
section,  the  outlets  being  located  back  of  transom  bar  on  approxi- 
mately 14  in.  (0.35  m.)  centers.  Each  lamp  is  equipped  with  a 
100-watt  size  concentrating  type  prismatic  reflector ;  the  units  are 
mounted  at  an  angle  so  that  tip  of  the  lamp  is  pointing  at  a  line  on 


e 


©  %  h  %  % 

"©  <5>©  '©  ©  ~@ 

O  B  ■  B  »  P 

®  ©  @  @  ©         <k 

©      %  %  %  ^4"-^ 

b  *®  @  ©^  © 

©  ©  ©  @  ®         © 

^@>        ^  ^  %  %>      ^ 


•     Ind/cates  /ocafo/i  of  Cc/</m"s. 

$ "  //ycf/cafes  /ocaf/0/7  of  £J$>S?/  Cfo/fe. 

■    facf/cofes  /ocof/on  of  Sfo/'/ons  oftrA/cA  ///i/mjnof/o/7  f?eaa'//7ffS  were  fafre/i 


Fig.  15. — Section  plan  of  third  floor  showing  location  of  illumination  test  stations. 

floor  approximately  two-thirds  of  the  distance  from  plate  glass 
to  opposite  enclosing  wall  of  window.  This  angle  is  found  to 
be  the  best  for  the  average  form  of  window  dress,  but  in  excep- 
tional cases,  provision  is  made  for  varying  the  angle.  The  light 
units  are  concealed  from  the  view  of  persons  standing  on  side- 
walk by  means  of  an  ornamental  drapery  placed  between  the 
light  units  and  plate  glass.    The  units  are  wired  so  that  alternate 


30  TRANSACTIONS  I.  E.  S. — PART  II 

lamps  are  on  the  same  circuit,  allowing  one  half  of  the  units  to  be 
used  on  ordinary  occasions  and  all  units  when  desired.  By  using 
a  larger  size  reflector  ioo-watt  lamps  may  be  used  on  special 
occasions  when  a  higher  intensity  of  illumination  is  required  in 
any  one  section  for  display  of  special  goods.  With  a  wattage 
allowance  of  but  4.7  per  square  foot  (0.09  sq.  m.),  very  satis- 
factory illumination  has  been  obtained.  It  will  be  noted  from 
fig.  10  that  a  very  small  amount  of  light  is  received  on  the  side- 
walk; practically  all  the  light  is  directed  on  the  goods  displayed. 

In  order  to  keep  the  light  from  the  upper  part  of  window  and 
also  from  interior  of  the  store,  a  metal  strip  has  been  placed  above 
the  light  units  and  painted  white  on  the  under  side.  A  small 
valance  has  been  suspended  from  this  strip  which  serves  to  cut 
off  the  light  from  the  interior  of  the  store.  The  fact  that  some 
light  is  received  directly  on  the  sidewalk  as  well  as  additional 
light  caused  by  direct  reflection  from  the  goods  on  display,  is 
worthy  of  attention,  since  such  light  serves  to  attract  the  atten- 
tion of  the  passersby. 

Show  Case  Illumination. — The  illumination  of  show  cases  in 
a  large  department  store  is  of  very  great  importance  and  a 
large  factor  in  the  use  of  energy.  In  the  present  instance  the  show 
cases  were  formerly  lighted  by  carbon  filament  tubular  type 
lamps,  but  are  now  lighted  far  more  satisfactorily  by  the  new 
tungsten  25-watt  drawn  wire  tubular  type  lamp  which  is  par- 
ticularly suited  for  this  class  of  lighting. 

NOTES  ON  THE  ILLUMINATION  TESTS. 

After  the  installation  was  made,  illumination  tests  were  con- 
ducted in  various  departments  to  determine  the  intensity  and 
uniformity  of  illumination  obtained. 

Test  No.  1  (First  floor;  buff  ceilings — soiled  condition;  6  years 
since  painted — buff  walls;  dark  wood  floor;  mahogany  fix- 
tures).— The  arrangement  of  the  furniture  made  it  necessary  to 
select  test  stations  as  shown  in  Fig.  13.  Horizontal  illumination 
readings  were  taken  on  a  plane  32  inches  (0.81  m.)  above  the 
floor,  which  is  the  height  of  the  counters  throughout  store.  The 
values  obtained  at  various  stations  are  shown  in  Fig.  13.  All 
values  were  corrected  for  any  changes  from  rated  voltage  of 
lamps  which  were  noted. 


shalling:    department  store  lighting 


31 


Test  No.  2  (Third  floor;  buff  ceilings — soiled;  walls  buff;  dark 
green  carpet;  mahogany  fixtures). — Test  stations  were  selected 
as  indicated  in  Fig.  15,  because  of  few  obstructions  in  this  bay. 
Horizontal  illumination  readings  were  taken  on  a  plane  32  in. 
(0.81  m.)  above  the  floor  or  at  the  counter  level.  The  following 
table  indicates  readings  obtained  at  various  stations : 


tation 

Foot-candles 

Station 

Foot-candles 

Station 

Foot-candles 

I 

1-93* 

15 

3-94 

29 

2.6o 

2 

2.09 

16 

4.42 

30 

2-93 

3 

2.12 

17 

3-95 

31 

2.76 

4 

2.1 1 

18 

3.62 

32 

2.61 

5 

2.28 

19 

4.08 

33 

2.76 

6 

2.30 

20 

4.40 

34 

3-oS 

7 

1-93* 

21 

3-94 

35 

2.60 

8 

2.84 

22 

3-63 

36 

1.89* 

9 

3-11 

23 

4.48 

37 

1.99 

10 

2-95 

24 

4-°3 

38 

2.01 

11 

2.77 

25 

3.62 

39 

2.09 

12 

2-93 

26 

4-13 

40 

2.13 

13 

3.01 

27 

4.12 

4i 

2.28 

14 

2.84 

28 

3-63 

42 

1.89* 

*  Indicates  light  partially  obstructed  by  column. 

Test  No.  3. — The  tests  were  conducted  on  another  floor  on 
which  the  change  from  arc  to  tungsten  lamps  had  not  yet  been 
made.  Tests  were  made  in  a  bay  occupying  the  same  relative 
position  as  bay  in  test  No.  2,  and  readings  were  made  at  the 
same  stations  under  practically  similar  conditions.  The  readings 
obtained  were  as  follows : 


Station 

Foot-candles 

Station 

Foot-candles 

Station 

Foot-candles 

I 

I.56* 

15 

2-33 

29 

I.98 

2 

I.62 

16 

2.80 

30 

2.07 

3 

1-55 

17 

2-93 

31 

i-95 

4 

I.5I 

18 

2.72 

32 

1-77 

5 

1-45 

19 

2.67 

33 

2.32 

6 

1.49 

20 

2.38 

34 

2.17 

7 

1-33* 

21 

2.28 

35 

1.58 

8 

1.92 

22 

2.36 

36 

1.56* 

9 

2.04 

23 

3.18 

37 

1.56 

10 

2.00 

24 

2.84 

38 

1.48 

11 

1.92 

25 

2.09  . 

39 

i-59 

12 

2.24 

26 

2-79 

40 

1.65 

13 

2.02 

27 

3-23 

4i 

1.58 

14 

i-95 

2S 

2.52 

42 

1.39* 

*  Indicates  light  partially  obstructed  by  column. 
Height  of  lamps  above  floor  10  ft.  6  in.  (3.20  m. ). 
Height  of  lamps  above  test  plane,  7  ft.  10  in.  (2.39  ni.). 
4 


32 


TRANSACTIONS  I.  E.  S. — PART  II 


Test  No.  4  (Eighth  floor;  buff  ceilings — soiled;  dark  green 
walls;  dark  floor;  mahogany  fixtures). — Horizontal  illumination 
readings  were  taken  on  this  floor  at  a  plane  32  in.  (0.81  m.) 
above  the  floor  at  similar  stations  and  under  similar  conditions 
as  Test  No.  2,  with  the  exception  that  four  light  units  are 
provided  per  bay,  each  consisting  of  a  60-watt  bowl-frosted 
tungsten  lamps  in  8  in.  light  opal  reflectors.1  The  results  of  test 
were  as  follows : 


Station 

Foot-candles 

Station 

Foot-candles 

Station 

Foot-candles 

I 

0.16* 

15 

1. 21 

29 

I.20 

2 

1. 21 

16 

1. 21 

30 

I.32 

3 

I-I5 

17 

i-i3 

31 

I.28 

4 

1. 17 

18 

1. 00 

32 

1. 19 

5 

1.15 

19 

1. 11 

33 

1. 21 

6 

1. 16 

20 

1. 16 

34 

1. 18 

7 

0.82* 

21 

1. 12 

35 

I- 13 

8 

1. 15 

22 

1. 19 

36 

O.86 

9 

1. 18 

23 

1.19 

37 

1.33 

10 

1.16 

24 

1.19 

38 

i-3t 

11 

1. 12 

25 

1. 11 

39 

1.26 

12 

1. 16 

26 

1.13 

40 

1.27 

13 

1. 17 

27 

1. 16 

4i 

1.20 

14 

1.09 

28 

1,15 

42 

1.03* 

*  Indicates  light  partly  obstructed  by  columns. 
Height  of  lamps  above  floor  approximately  10  ft.  6  in.  (3.20  m.). 
Height  of  lamps  above  test  plane  approximately  7  ft.  10  in.  (2.39  m. ). 

In  all  the  foregoing  tests  it  was  necessary  to  choose  bays  for 
test  purposes  which  were  practically  free  from  obstruction,  and 
while  the  readings  obtained  for  any  one  bay  if  properly  con- 
sidered will  give  a  fair  idea  of  the  average  intensity  for  that 
bay,  these  figures  have  not  been  given  inasmuch  as  they  might 
be  construed  as  representing  average  intensity  for  whole  floor 
space. 

CLEANING  AND  MAINTENANCE  OF  PRESENT  INSTALLATION. 
A  careful  system  for  regular  inspection  and  cleaning  has  been 
initiated  in  this  store  whereby  all  units  in  the  building  are 
brushed  off  with  a  stiff  brush  and  cloth  once  every1  two  weeks, 
and  each  unit  is  taken  down  and  washed  thoroughly  once  each 
month.  The  cost  of  cleaning,  including  labor  and  materials,  is 
approximately    $350   per   year    for    approximately    1,000    units 

1  Holophane  "  Veluria." 


shalling:    department  store  lighting  33 

installed.  Since  the  units  are  thoroughly  cleaned  once  per 
month,  the  cost  of  cleaning  per  unit  will  be  approximately 
3  cents,  which  includes  one  dusting  and  one  washing.  A  fair 
division  of  this  cost  would  be  y2  cent  per  unit  for  dusting  and 
2]/2  cents  for  thorough  cleaning.  Under  such  a  system  the 
illuminating  efficiency  is  kept  at  a  practical  maximum.  It  is 
probable,  however,  that  in  other  places  where  dirt  and  dust  con- 
ditions are  not  as  severe,  that  the  cleaning  could  be  done  at  less 
frequent  intervals.  The  above-given  figures  represent  a  very 
small  part  of  the  operating  expense,  and  are  considerably  less 
than  cost  of  trimming  and  other  maintenance  charges  incident 
to  the  operation  of  the  old  system. 

CONCLUSION. 

With  the  aforementioned  changes,  this  store  now  has  a  thor- 
oughly modern  and  attractive  lighting  equipment,  a  considerably 
higher  intensity  of  illumination  at  a  saving  in  operating  expense 
of  over  50  per  cent.  The  figures  given  for  cleaning  and  main- 
tenance serve  to  indicate  what  may  be  accomplished  by  syste- 
matic attention  to  this  detail  at  a  cost  probably  not  as  great  as 
the  cost  of  cleaning  the  windows. 

The  writer  desires  to  take  this  opportunity  of  heartily  thank- 
ing Messrs.  L.  J.  Kiefer,  of  the  McCreery  Company,  and  P.  C. 
Keller  for  the  valuable  assistance  which  they  rendered  him  in 
collecting  the  data  for  this  paper. 

DISCUSSION. 

Mr.  G.  H.  Stickney  :  This  paper  contributes  some  important 
data  on  the  recent  practise  in  department  store  or  rather,  as  I 
should  class  McCreery's  Store,  dry  goods  store  lighting.  It  is, 
therefore,  of  great  value  to  all  those  who  have  lighting  problems 
of  this  character. 

A  large  dry  goods  or  department  store  combines  in  one  institu- 
tion a  large  number  of  departments  or  stores.  As  such,  each 
department  has  its  own  lighting  requirements  pertaining  to  the 
class  of  merchandise  handled.  Thus,  we  find  the  problems  of 
the  silk  store,  furniture  store,  picture  store,  jewelry  store, 
restaurant,  etc.,  but  in  addition,  since  these  are  all  brought 
together  in  one  establishment,  the  lighting  of  each  must  bear  a 


34  TRANSACTIONS  I.  E.  S. — PART  II 

relation  to  all  the  others  in  order  that  we  may  have  a  unity  and 
dignity  in  keeping  with  the  large  institution. 

Referring  to  some  detail  points  in  the  paper,  I  note  that  the 
author  believes  that  the  brightness  of  the  lighting  units  should  be 
equal  to  that  of  the  ceiling.  From  my  own  observation,  I  am  led 
to  the  opinion  that,  for  appearance  sake,  it  is  desirable  in  an  in- 
stallation of  this  type  to  have  the  units  preferably  brighter  than 
the  ceiling,  although  this  should  not,  of  course,  be  carried  far 
enough  to  introduce  objectionable  glare. 

As  to  color  values,  there  has,  in  the  past,  been  considerable 
difference  of  opinion  regarding  the  requirements  of  dry  goods 
stores  and  the  practise  in  different  cities  has  differed.  There  are 
certain  departments,  such  as  dress  goods  and  silk  departments, 
where  accurate  color  selection  is  desirable;  both  daylight  and  the 
prevailing  evening  light  should  be  available  so  that  goods  for 
street  wear  may  be  selected  by  daylight  and  those  for  evening 
wear  by  artificial  light.  One  condition  which  is  often  overlooked 
is  that  a  high  intensity,  say  20  foot-candles  or  more,  is  necessary 
for  accurate  daylight  selection  of  delicately  colored  materials. 
Moreover,  the  only  accurate  color  matching  artificial  lights  are 
all  relatively  inefficient  and,  as  a  result,  it  is  not  practicable,  in 
the  present  state  of  the  art,  to  light  an  entire  department  by 
artificial  light  for  daylight  color  matching.  The  best  solution  of 
this  problem  is  that  adopted  by  the  McCreery  Company,  namely, 
that  of  providing  small  rooms  or  booths  where  artificial  daylight 
of  the  proper  diffusion  and  intensity  can  be  obtained  in  the  even- 
ing, and  correspondingly,  evening  light  in  the  daytime.  Even 
where  color  matching  booths  are  not  provided,  nearly  all  the  large 
dry  goods  and  department  stores  are  turning  to  the  tungsten  fila- 
ment lamp.  Among  these  may  be  mentioned  Marshall  Field 
Co.,  Carson  Piru  Scott  &  Co.,  Mandel  Brothers,  Rothchilds  and 
The  Fair  in  Chicago,  and  the  McCreery  Co.,  Gimbels  and  Green- 
hut-Siegel-Cooper  Co.,  in  New  York.  There  are  many  others 
nearly,  if  not  equally  prominent. 

As  expressed  to  me  by  the  manager  of  one  of  the  large  dry 
goods  stores  in  Chicago,  the  pleasing  appearance  of  his  store 
under  the  tungsten  filament  light  much  more  than  compensated 
for  the  color  matching  advantages  of  any  other  light.     There  are 


shaujng:     department  store  lighting  35 

relatively  few  departments  in  such  a  store  where  accurate  color 
matching  is  necessary,  while  in  other  departments,  the  warm 
color  tone  is  considered  so  important  that  it  is  often  the  practise 
to  exclude  daylight  by  mean  of  shades  or  screens  and  display  the 
goods  throughout  the  day  by  the  warm  light  from  the  tungsten 
tilament  lamp. 

The  enterprise  of  this  company  in  making  the  new  installation 
without  in  any  way  interfering  with  the  sales  activity  of  the  store 
is  commendable.  I  know  that  this  meant  considerable  planning 
beforehand.  In  many  stores  even  the  simple  maintenance  is  not 
planned  so  as  to  avoid  such  interference. 

Another  point  worthy  of  mention  is  the  excellent  clean  con- 
dition in  which  we  found  the  installation  in  our  inspection  to-night. 
YVe  cannot  emphasize  too  much  the  importance  of  cleanliness  in 
maintaining  the  efficiency  and  appearance  of  any  lighting  system. 
The  practise  of  regular  inspection  and  cleaning  at  prescribed 
intervals  is  the  only  way  to  insure  this  result,  and,  as  brought  out 
in  the  paper,  the  cost,  even  in  Pittsburgh,  is  surprisingly  low. 

The  whole  installation  is  admirably  adapted  for  its  purpose  and 
is  one  of  the  best  I  have  seen  anywhere. 

Mr.  W.  M.  Skiee :  With  reference  to  showcase  lighting,  I 
would  like  to  call  attention  to  the  reduction  in  heat  in  showcases 
due  to  the  use  of  tungsten  filament  lamps.  With  a  50  watt 
tubular  gem  filament  lamp  the  glass  on  showcases  is  often  heated 
and  the  breakage  of  the  glass,  due  to  expansion,  is  common. 
With  the  tungsten  filament  lamp  of  half  the  wattage  28  per  cent, 
more  candle-power  is  obtained  than  with  the  carbon  lamp,  and 
there  is  less  liability  of  breakage  of  the  glass  showcase  due  to 
heat. 

In  connection  with  the  color  of  light,  my  attention  was  recently 
called  to  a  theatrical  company  which  made  a  practise  of  carrying 
an  equipment  of  strip  and  border  lights-  composed  of  32  candle- 
power  carbon  filament  incandescent  lamps,  part  of  which  were 
amber  dipped.  This  was  departing  to  a  considerable  extent  from 
daylight  values  and  the  reason  given  was  that  the  make-ups  were 
artificial  and,  therefore,  required  other  than  daylight  values  to 
produce  the  best  results.     It  is  no  doubt  true  that  daylight  value 


36  TRANSACTIONS  I.  E.  S. — PART  II 

in  stores  is  not  necessary  for  the  proper  illumination  of  materials 
which  will  be  used  in  the  home  or  elsewhere  under  incandescent 
electric  lamps. 

Mr.  S.  G.  Hibben  :  With  reference  to  the  question  of  main- 
tenance, it  would  seem  that  in  retail  sales  houses  where  the  clean- 
ing is  not  as  thorough  or  does  not  follow  as  definite  a  schedule  as 
in  this  store,  that  of  reflectors  of  equal  efficiency  the  one  which 
will  show  dirt,  and  consequently  the  need  of  proper  cleaning  and 
up-keep,  is  preferable. 

On  going  through  the  store  I  also  noticed  considerable  glare 
from  the  upper  parts  of  the  prismatic  units,  and  it  is  suggested 
that  frosting  of  these  upper  portions  would  be  highly  advisable. 
A  commendable  point  I  wish  to  call  attention  to  is  the  fact  that 
the  new  units  on  most  of  the  floors  are  placed  higher  than  were 
the  replaced  arc  lamps.  In  the  low  position,  the  arc  lamps  were 
decidedly  unpleasant. 

On  the  page  referring  to  window  lighting  it  is  stated  that  a 
large  reflector  is  used  so  that  small  or  large  lamps  may  be  em- 
ployed as  desired.  I  would  like  to  ask  if  it  is  not  true  that  in  this 
case,  where  the  small  lamp  is  used  in  the  reflector  designed  for 
the  next  larger  sized  lamp,  that  the  candle-power  distribution 
is  not  changed,  since  the  filament  position  is  different  than  that 
for  which  the  reflector  was  originally  designed. 

I  note  also  that  on  different  floors  the  same  sized  glass 
accessories  are  used,  but  different  sized  lamps  are  employed.  I 
believe  the  author  has  neglected  to  mention,  in  this  connection, 
the  very  desirable  feature  of  these  fixtures  is  that  there  is  pro- 
vision made  for  a  mechanical  adjustment  to  allow  of  nearly  the 
same  distribution  when  using  various  sizes  of  lamps. 

It  must  be  remembered  that  the  paper  cannot  be  used  as  a  basis 
of  design,  since  it  is  essentially  a  description  of  an  installation. 
Some  of  the  main  units  employed  are  admittedly  not  the  most 
efficient  that  might  have  been  obtained.  They  were  not  selected 
on  account  of  color  values-;  their  maintenance  cost  is  not  low;  nor 
is  their  deterioration  small.  It  is  said  that  their  harmonizing  and 
decorative  appearance  determined  their  selection,  and  of  course 
when  one  discusses  taste  there  is  a  wide  latitude  of  opinion. 


shaixing:    department  store  lighting  37 

Mr,  Ward  Harrison  :  With  reference  to  maintenance,  we 
have  found  that  with  open  reflector  equipment  approximately 
one-half  of  the  deterioration  in  the  illuminating  value  of  the 
reflectors  arises  from  the  deposit  of  dust  on  the  outside  of  the 
reflectors,  and  one-half  is  due  to  dust  on  the  inside  and  on  the 
lamp  bulb.  When  the  lamp  is  enclosed  as  in  the  McCreery  fix- 
tures, one-half  of  this  deterioration  is  largely  eliminated.  This, 
with  the  fact  that  a  clear  instead  of  a  bowl-frosted  lamp  is  used, 
makes  up  in  a  large  measure  for  the  lower  efficiency  to  be 
expected  because  of  enclosing  the  lamp. 

The  complete  change  of  the  lighting  equipment  on  a  floor  in 
one  night  is  a  commendable  plan  because  of  the  fact  that  the 
contrast  in  the  colors  of  the  arc  and  incandescent  lamps  is  not 
pleasant  and  often  store  employees  are  not  in  favor  of  the  change 
when  they  see  the  two  systems  thus  compared.  Both  illuminants 
vary  from  daylight ;  one  has  an  excess  of  red,  the  other  of  blue 
rays ;  and  in  juxtaposition  the  effect  is  not  pleasing. 

In  passing  through  the  store  I  noticed  that  the  ceilings  were 
of  such  a  color  that  their  coefficient  of  reflection  was  not  high. 
Had  they  been  painted  white,  the  efficiency  of  the  installation 
would  have  been  increased  appreciably  and  the  contrast  in  bright- 
ness between  light  sources  and  the  ceiling  would  also  have  become 
less  marked. 

With  reference  to  the  illumination  tests.  I  would  like  to  ask 
Mr.  Shalling  whether  the  foot-candle  values  given  are  in  each 
case  representative  of  the  whole  floor  area,  and  also  at  what 
period  during  the  life  of  the  lamps  the  tests  were  conducted;  in 
a  word,  whether  the  data  as  given  are  representative  of  average 
conditions  including  depreciation  due  to  dust. 

Mr.  L.  J.  Kiefer  :  The  old  equipment  consisted  of  standard 
enclosed  carbon  arc  lamps,  not  intensified  arc  lamps. 

The  heat  formerly  given  off  by  carbon  lamps  in  showcases 
where  jewelry  was  displayed  was  such  that  injury  frequently 
resulted  to  jewel  settings,  so  much  so  that  it  was  a  common  thing 
to  send  material  back  to  the  factory  to  have  the  stones  reset.  The 
use  of  high  efficiency  tungsten  lamps  has  eliminated  this  trouble. 

The  question  of  color  values  has  been  carefully  considered  in 
connection  with  the  lighting  of  the  McCreery  &  Company  store. 


38  TRANSACTIONS  I.  E.  S. — PART  II 

On  the  second  floor  there  is  provided  convenient  space  next  to 
the  windows  where  clerks  may  show  silks  and  dress  goods  under 
daylight.  For  dark  days  and  during  a  few  hours  in  the  evening 
in  some  seasons,  there  is  also  a  special  room  equipped  so  that  a 
customer  may  see  goods  under  both  evening  and  daylight  con- 
ditions. An  intensified  arc  with  a  color  screen  illuminates  the 
room,  forming  a  circular  panel  of  white  light  on  the  ceiling. 
Incandescent  lighting  is  also  provided  and  can  be  conveniently 
switched  on  as  a  customer  desires. 

Mr.  H.  W.  Shaixing  (in  reply)  :  The  units  formerly  used  in 
this  store  were  not  the  intensified  type  of  arc  lamps,  but  were 
standard  enclosed  carbon  arc  lamps. 

In  designing  the  installation,  it  was  found  by  careful  experi- 
ments that  satisfactory  color  matching  for  most  practical  pur- 
poses could  be  obtained  by  the  use  of  the  tungsten  filament  lamp. 
In  some  cases  a  much  closer  approximation  to  true  daylight 
values  was  obtained  by  burning  the  lamps  above  their  rated 
voltage.  For  particular  cases,  daylight  can  be  used  as  a  general 
rule  by  carrying  goods  to  the  windows ;  a  booth  is  also  available 
with  a  specially  prepared  color  matching  device. 

In  answer  to  Mr.  Harrison's  questions,  I  would  say  that  the 
areas  chosen  for  test  purposes  were  representative;  all  lamps  on 
any  one  floor  were  burning  during  the  tests.  Slightly  different 
intensities  might  be  obtained  near  the  windows  or  side  walls,  but 
an  average  of  the  values  found  can  be  considered  to  be  practically 
an  average  for  the  entire  floor  space.  The  units  had  been  in  use 
about  three  months  when  tests  were  made. 

In  the  cut  glass  department,  clear  reflectors  and  clear  lamps  are 
used.  In  the  showcases  a  metal  reflector  having  a  polished  sur- 
face is  used.  In  the  millinery  showcases,  polished  metal 
reflectors  are  also  used. 

In  answer  to  Mr.  Hibben's  remarks,  it  will  be  noted  that  in  the 
clothing  department  where  400  watt  units  are  employed  with 
lamps  burned  above  their  rated  voltage,  that  excellent  color 
values  are  obtained  while  the  glare  is  not  at  all  objectionable  or 
even  noticeable  to  the  average  customer,  considering  the  time 
such  customer  would  usually  spend  in  this  department. 

On  the  first  floor  in  order  to  obtain  the  required  intensity  and 


shaeung:     department  store  lighting  39 

to  get  the  desired  ornamental  effects,  it  was  necessary  to  use 
about  twice  the  wattage  ordinarily  used  on  the  other  floors. 

In  designing  an  installation  for  a  store  of  this  character  it  is 
necessary  to  carefully  analyze  conditions  on  the  various  floors  in 
order  that  there  will  not  be  too  great  a  sameness  to  the  units 
employed.  The  units  should  be  in  harmony  with  the  architectural 
and  fixture  conditions. 


40  TRANSACTIONS  I.  E.  S. PART  II 

TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  UNDER 

DIFFERENT  SYSTEMS  OF  ILLUMINATION  AND 

A  PRELIMINARY  STUDY  OF  THE  CAUSES 

OF  DISCOMFORT.* 


BY  C.  E.  FERREE. 


Synopsis: — Besides  outlining  (I)  the  problem  which  confronts  the 
investigator  who  would  determine  the  effects  of  various  lighting  systems 
on  the  eye,  this  paper  discusses:  (II)  the  scale  or  general  level  of  effi- 
ciency of  the  eye  under  different  systems  of  lighting,  with  brief  comments 
on  the  conventional  tests  for  the  efficiency  of  the  eye  such  as,  (a)  color 
discrimination,  (&)  brightness  discrimination,  (c)  visual  acuity — the  latter 
tests,  modified,  it  is  contended  are  adequate  for  the  determination  of  the 
general  level  of  efficiency  of  the  unfatigued  eye;  (III)  loss  of  visual 
efficiency  as  the  result  of  a  period  of  work — here  it  is  contended  that 
each  of  the  aforementioned  tests  fails  to  show  a  true  loss  of  visual  effi- 
ciency, and  a  new  test  is  described.  The  paper  is  concluded  (IV)  with  a 
brief  statement  of  some  of  the  causes  of  ocular  discomfort  under  various 
conditions,  and  a  description  of  a  method  of  making  a  comparative  esti- 
mate of  discomfort. 

I.  INTRODUCTION. 

In  191 1  the  American  Medical  Association  appointed  a  com- 
mittee to  study  the  effect  of  different  lighting  systems  on  the  eye. 
The  writer  was  asked  to  share  in  the  work  of  this  committee. 
The  problem  presented  to  him  was  to  furnish  tests  that  would 
show  the  effect  of  different  lighting  systems  on  the  eye  and  more 
especially  to  devise,  if  possible,  a  test  that  would  show  a  loss  of 
efficiency  as  the  result  of  three  or  four  hours  of  work  under  an 
unfavorable  lighting  system.  It  is  the  purpose  of  the  following 
paper  to  give  a  preliminary  report  of  the  work  that  has  been 
carried  on  by  the  writer  in  this  field  during  the  past  year. 

Confronting  the  problem  of  the  effect  of  lighting  systems  on 
the  eye,  it  is  obvious  that  the  first  step  toward  systematic  work 
is  to  obtain  some  means  of  making  a  definite  estimate  of  this 
effect.  The  prominent  effects  of  bad  systems  of  lighting  are  loss 
of  efficiency,  temporary  and  progressive,  and  eye  discomfort. 
Having  devised  methods  which  after  six  months  of  testing  he 
has  found  to  be  accurate  and  practicable,  the  writer  has  under- 

*  A  paper  read  at  the  sixth  annual  convention  of  the  Illuminating  Engineering 
Society,  Niagara  Falls,  Ont.,  September  16-19,  J912- 


ferree:    tests  for  the  efficiency  of  the  eve  41 

taken  to  determine  (1)  the  lighting  conditions  that  give  in  gen- 
eral the  highest  level  or  scale  of  visual  efficiency,  (2)  the  condi- 
tions that  give  the  least  loss  of  efficiency  for  continued  work, 
and  (3)  the  conditions  that  cause  the  least  discomfort.  This 
plan  of  work,  it  is  scarcely  needful  to  remark,  will  involve  a 
wide  range  of  experimentation.  The  crux  of  the  problem,  as  the 
writer  conceives  it,  is,  however,  to  secure  reliable  methods  of 
estimating  effect.  Having  these  methods,  the  factors  whatever 
they  may  be,  intensity,  quality,  position  of  light  relative  to  the 
eye,  etc.,  can  be  varied  one  at  a  time  and  the  effects  be  determined. 
From  these  effects  it  should  not  be  difficult  to  ascertain  what 
lighting  conditions  are  best  for  the  eye  and  what  is  the  relative 
importance  of  the  factors  that  go  to  make  up  these  conditions. 
Further  it  should  be  possible  on  the  practical  side  to  test  out  and 
perfect  a  lighting  system,  so  far  as  its  effect  on  the  eye  is  con- 
cerned, before  we  put  it  on  the  market.* 

In  this  report  nothing  more  will  be  attempted  than  to  indicate 
what  methods  may  be  used  in  the  three  steps  of  the  problem  as 
outlined  above. 

II.  THE    SCALE   OR  GENERAL    LEVEL    OF    EFFICIENCY  OF 
THE  EYE  UNDER  DIFFERENT  SYSTEMS  OF  LIGHTING. 

A  general  survey  of  the  field  shows  that  at  different  times  the 
following  tests  have  been  used  in  one  capacity  or  another  for  de- 
termining the  efficiency  of  the  eye :  brightness  discrimination,  col- 
or discrimination,  and  visual  acuity.  No  extensive  use,  if  any  at 
all,  has  been  made  of  any  of  these  with  the  exception  of  visual 
acuity  in  connection  with  problems  of  the  type  here  considered, 
but  the  fitness  of  their  application  in  some  form  to  such  problems 
is  evident  at  a  glance.  If  the  eye's  efficiency  is  to  change  at  differ- 
ent times  and  under  different  conditions  of  lighting,  it  should  be 
manifested  in  changes  in  brightness  discrimination,  color  discrimi- 
nation, or  visual  acuity.  The  first  step  in  our  work  would,  then, 
seem  to  be  to  devise  for  these  points  tests  which  are  sufficiently 
sensitive  for  use  in  work  of  the  kind  we  have  in  hand.  The  gen- 
eral nature  of  these  tests  is  too  familiar  to  need  detailed  mention 
here.  A  few  special  points  may.  however,  be  given  in  passing. 
(1)  The  threshold  or  limen  test  is  the  most  sensitive  and  practical 

*  This  latter  point  was  suggested  to  the  writer  by  reading  Dr.  Ives'  discussion  of  this 
paper  (p.  57). 


42  TRANSACTIONS  I.  E.  S. — PART  II 

for  color  sensitivity.  In  making  this  test  the  pre-exposure1  and 
the  surrounding  field2  should  be  of  a  gray  of  the  brightness  of 
the  color  at  or  near  its  threshold  value.  Further,  the  illumination 
of  the  room  must  be  kept  constant  from  test  to  test.3  If  the 
colored  light  is  to  be  obtained  by  reflection,  disks  of  standard 

1  By  pre-exposure  is  meant  what  the  eye  rests  on  immediately  preceding  its  stimula- 
tion by  color.  It  is  obvious  that  there  must  always  be  some  pre-exposure  and,  unless 
care  be  taken  to  eliminate  its  effect,  it  will  influence  the  eye's  sensitivity  to  color.  Even 
closing  the  eye,  as  is  often  done  before  stimulating  by  color,  is  the  equivalent  of  giving  a 
black  pre-exposure.  All  color  must  of  course  be  eliminated  from  the  pre-exposure.  It 
should  also  be  of  the  same  brightness  as  the  color  by  which  the  e3re  is  to  be  stimulated.  If 
not  it  gives  a  brightness  after-image  which  mixes  with  the  succeeding  color  impression 
and  reduces  its  saturation.  This  reduction  of  saturation  takes  place  apparently  at  some 
physiological  level  posterior  to  the  seat  of  the  positive,  negative,  and  contrast  color 
processes  commonly  supposed  to  be  located  in  the  retina.  (See  Ferree  and  Rand  :  "  The 
Fusion  of  Brightness  with  Color — The  I,ocus  of  the  Action,"  Journal  of  Philosophy,  Psy- 
chology and  Scientific  Methods,  VIII,  1911^.294.)  If  the  pre-exposure  is  lighter  than  the 
coior  it  adds  \>y  after-image  a  certain  amount  of  black  to  the  succeeding  color  impression 
and,  if  darker,  it  adds  a  certain  amount  of  white.  Since  white  inhibits  color  more  than 
black,  the  effect  of  a  dark  pre-exposure  is  to  reduce  the  sensitivity  to  color  more  than  the 
effect  of  a  light  pre-exposure.  But  since  both  white  and  black  as  after-effect  reduce  the 
sensitivity  to  color,  the  eye  is  rendered  more  sensitive  when  no  after-image  is  given,  i.  e. 
when  the  pre-exposure  is  of  the  same  brightness  as  the  color.  The  pre-exposure  therefore 
should  be  to  a  gray  of  the  brightness  of  the  color.  No  brightness  after-image  will  be 
added  thereby  to  the  succeeding  color  impression  to  modify  either  its  saturation  or  color 
tone. 

2  When  the  surrounding  field  is  either  lighter  or  darker  than  the  color,  brightness  is 
induced  by  contrast  across  the  colored  surface.  When  the  surrounding  field  is  lighter 
than  the  color,  a  certain  amount  of  black  is  induced,  and  when  darker,  a  certain  amount 
of  white  is  induced.  As  stated  above,  the  mixture  of  this  white  or  black  with  color, 
although  it  does  not  alter  the  amount  of  colored  light  coming  to  the  eye.  reduces  the 
saturation  of  the  color.  The  effect  of  brightness  contrast  can  be  eliminated  only  by 
making  the  brightness  of  the  surrounding  field  a  gray  of  the  brightness  of  the  color. 
This  can  be  done  by  means  of  a  gray  screen  around  the  color,  or  by  a  larger  gray  disk  in 
case  a  color  mixer  is  used. 

3  In  case  the  colored  light  used  for  the  stimulus  is  obtained  by  reflection  from  a 
pigment  surface,  a  change  in  the  general  illumination  of  the  field  of  vision  affects  the 
results  of  the  sensitivity  tests  in  the  following  ways.  (1)  It  changes  the  amount  of 
colored  light  coming  to  the  eye.  (2)  By  changing  its  brightness  adaptation  it  changes 
the  sensitivity  of  that  part  of  the  retina  upon  which  the  colored  light  falls.  (3)  By 
changing  the  sensitivity  of  the  eye  to  brightness  after-image  and  contrast,  it  changes  the 
amount  of  brightness  added  to  the  color  as  the  result  of  pre-exposure  and  surrounding 
field,  and  therefore  changes  the  effect  of  pre-exposure  and  surrounding  field  upon  the  color 
impression.  Moreover,  the  effect  of  pre-exposure  and  surrounding  field  cannot  be  elimin- 
ated even  when  both  are  made  of  the  brightness  of  the  color  for  some  given  illumination, 
unless  that  illumination  be  kept  constant  throughout  the  test  for,  when  it  changes,  the 
brightness  of  the  color  and  of  the  grays  used  as  pre-exposure  and  surrounding  field  does 
not  change  in  equal  amounts  ;  hence,  the  brightness  equality  which  is  needed  cannot  be 
maintained.  In  case  the  colored  light  is  not  gotten  by  reflection  from  a  pigment  surface 
but  is  obtained  from  monochromatic  sources  from  standard  filters  or  from  the  spectrum, 
only  the  last  two  of  the  factors  stated  above  influence  the  results  of  the  tests  for  color 
sensitivity.  In  the  tests  made  by  the  writer,  the  general  illumination  was  rendered 
constant  by  methods  to  be  described  later  in  the  paper. 

Although  for  the  purposes  of  this  work  the  tests  for  color  sensitivity  could  never  be 
conducted  in  the  dark-roam,  still  it  may  be  of  general  interest  to  note  at  this  point  that  the 
elimination  of  the  effect  ot  pre-exposure  and  surrounding  field  cannot  be  accomplished 
in  work  on  color  sensitivity  done  in  the  dark-room,  because  in  the  dark-room  the  pre- 
exposure and  surrounding  field  cannot  be  made  of  the  brightness  of  the  color.  They 
will  always  therefore  exert  an  effect  on  the  color  impression.  Moreover  since  the  colors 
all  differ  in  brightness,  this  effect  will  be  exerted  in  different  amounts  on  the  different 
colors.  That  is,  the  amount  of  brightness  added  by  after-image  or  contrast  depends 
upon  the  amount  of  brightness  difference,  respectively,  between  pre-exposure  and  color 
and  surrounding  field  and  color.  As  stated  above  this  amount,  when  working  in  the 
dark-room,  will  be  different  for  the  different  colors.  For  this  reason  and  also  because 
even  the  same  amount  of  brightness  excitation  acts  with  different  degrees  of  strength 
upon  the  excitation  set  up  by  the  different  colors,  it  is  especially  important  that  110  work 
on  the  comparative  sensitivity  of  the  retina  to  the  different  colors  should  be  done  in  the 
dark-room.  It  should  be  done  in  a  light  room  of  a  constant  intensity  of  illumination  and 
with  pre-exposure  and  surrounding  field  in  each  case  of  the  brightness  of  the  color  to  be 
used.  In  this  way  alone  can  all  the  factors  which  influence  the  sensitivit}'  of  the  retina, 
extraneous  to  the  source  of  light,  be  eliminated. 


FERREE:      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  43 

colored  and  gray  papers  (e.  g.,  the  papers  of  the  Hering  series) 
may  be  used  on  a  color  mixer.4  If,  on  the  other  hand,  it  is 
desirable  to  use  the  light  of  the  spectrum  or  the  light  transmitted 
through  standard  niters,  the  colored  light  may  be  cut  down  to 
the  threshold  value  by  means  of  a  sectored  disk,  the  sectors  of 
which  should  be  covered  with  a  gray  of  the  brightness  of  the 
color  at  or  near  its  threshold  value.  (2)  For  brightness  discrimi- 
nation also  the  threshold  or  limen  test  is  the  most  sensitive  and 
practical,  but  when  made  in  a  well-illuminated  room,  it  becomes 
in  effect  a  test  for  a  just  noticeable  difference.  This  test  may  be 
performed  at  different  points  in  the  brightness  scale,  e.  g.,  when 
the  standard  is  black,  near  mid-gray,  or  white.  As  before,  disks 
of  standard  papers  may  be  used  on  the  color-mixer,  or  the  light 
from  a  given  source  may  be  varied  by  means  of  a  sectored  disk.5 
(3)  Visual  acuity  tests  of  the  Snellen  type,  especially  when  used 
in  work  in  which  it  is  required  to  make  successive  tests  on  the  same 
person,  are  open  to  the  following  objections,  (a)  The  judgment 
is  in  terms  of  recognition.  A  letter  may  be  recognized  when  it  is 
not  seen  clearly.  In  any  judgment  based  on  the  recognition  of 
even  a  single  letter,  memory  plays  an  important  role.  It  is,  so  far 
as  the  writer  knows,  impossible  to  standardize  this  memory  fac- 
tor and  to  obtain  results  strictly  in  terms  of  acuteness  of  vision. 
(b)  The  test  card  is  made  up  of  quite  a  long  series  of  letters. 
As  the  test  progresses  the  letters  are  memorized  more  and  more 
completely.  It  is  practically  impossible  to  eliminate  this  progres- 
sive error  when  a  number  of  successive  judgments  have  to  be 

4  In  making  the  tests  with  reflected  light,  two  sets  of  disks  are  mounted  on  a  color 
mixer  (a)  an  outer  disk  of  gray  of  the  brightness  of  the  color  to  be  used,  and  (b)  an  inner 
disk  made  up  of  this  gray  and  the  disk  of  color.  To  the  inner  gray  disk,  varying  pre- 
portions  of  the  color  are  added  until  the  threshold  value  or  just  noticeable  color  is 
obtained.  To  facilitate  the  judgment  of  just  noticeable  color,  the  inner  disk  of  gray  plus 
color  is  compared  with  the  outer  disk  of  gray  as  a  standard.  Since  both  grays  are  of  the 
brightness  of  the  color,  the  addition  of  the  colored  sector  to  the  inner  disk  produces  no 
change  of  brightness  either  to  confuse  the  judgment  of  noticeable  color,  or  to  affect  the 
intensity  of  the  color  excitation  actually  aroused.  In  getting  the  threshold  value,  the 
method  of  ascending  and  descending  series  should  be  used,  that  is,  beginning  with  equali- 
ty, the  variation  is  towards  noticeable  difference  and  beginning  with  a  difference  greater 
than  noticeable,  the  variation  is  towards  equality.  An  average  of  the  two  sets  of 
results  is  taken  for  the  threshold  value. 

5  When  the  test  is  made  with  reflected  light  two  sets  of  disks,  an  outer  and  an  inner, 
are  mounted  on  the  color  mixer.  Each  set  is  made  up  of  one  white  and  one  black  disk. 
Both  sets  of  disks  are  set  at  the  point  in  the  brightness  scale  from  which  the  variation 
towards  white  or  black  is  to  be  made.  One  is  kept  constant  and  the  other  is  changed 
until  the  judgment  different  is  given.  In  making  the  judgment  the  method  of  ascending 
and  decending  series  is  used  and  the  results  are  averaged  for  the  difference  limen.  This 
difference  limen  is  taken  as  the  measure  of  the  observer's  sensitivity  to  brightness  or 
white  light. 


44  TRANSACTIONS  I.  E.  S. PART  II 

made  as  is  the  case  before  a  final  result  is  reached  in  any  single 
visual  acuity  test  and  as  is  especially  the  case  when  a  number  of 
successive  tests  have  to  be  given  to  the  same  person,  which 
happens  in  much  of  the  work  involved  in  the  solution  of  the 
problem  here  proposed.  It  might  be  supposed  that  the  memoriza- 
tion of  the  series  could  be  broken  up  by  using  in  each  successive 
judgment  in  a  single  test  or  in  the  successive  tests,  as  the  case 
may  be,  cards  having  a  different  distribution  of  the  letters  in  the 
series.  Considerable  inconvenience  would,  however,  be  involved 
in  giving  the  tests  in  this  way  and  besides  no  guarantee  could  be 
had  that  each  judgment  would  present  the  same  degree  of  diffi- 
culty. That  is,  the  series  is  made  up  of  similar  and  dissimilar 
letters.  The  dissimilar  letters  can  be  distinguished  from  each 
other  with  less  difficulty  than  the  similar.  It  is  practically  impos- 
sible to  distribute  the  letters  so  that  the  individual  tests  may  be 
equally  rigorous.  This  objection  can,  of  course,  be  eliminated 
in  part  by  a  careful  selection  of  the  test  letters,  but  not  entirely 
because  a  series  of  letters  uniformly  similar  cannot  be  found, 
(c)  The  Snellen  series  contains  quite  a  large  number  of  letters. 
The  eye  is  found  to  fatigue  and  vision  to  blur  before  the  series 
is  completed.  This  introduces  an  error  which  it  is  practically 
impossible  to  render  constant. 

All  of  the  above  objections  were  eliminated  in  the  tests  finally 
adopted  by  us  by  changing  the  type  of  judgment  and  by  making 
the  test  object  in  one  case  two  parallel  vertical  lines  stamped  I 
mm.  apart  on  a  white  card6  and  in  another  the  letters  li  printed 
in  small  type.7     In  using  these  cards  the  observer's  acuity  of 

6  The  card  is  mounted  on  a  sliding  carrier  which  runs  on  a  track  made  of  two  meter 
rods  fastened  end  to  end  on  a  folding  base.  The  base  is  mounted  on  adjustable  stands 
fastened  to  a  table.  When  making  the  test  the  apparatus  is  so  adjusted  that  the  track 
carrying  the  test  card  is  just  below  and  close  to  the  observer's  eye.  In  order  that  the 
observer's  head  maybe  held  steady  he  is  required  to  bite  an  impression  of  his  teeth, 
previously  made  and  hardened  in  wax  on  a  mouth-board,  which  is  rigidly  fastened  by  a 
heavy  iron  rod  and  accessories  to  the.table  supporting  the  track  and  carrier. 

7  Besides  the  letters  lithe  writer  would  recommend  the  following  figures  as  test  objects. 

I 
(i)  — ■ — .    The  test  is  to  distinguish  clearlv  the  dot  at  the  center.    A  test  object  in  the 

I 
shape  of  a  cross  has  the   advantage  of  affording  a  steady  control  of   fixation.     According 
to   photographic   records   of    involuntan7  eye-movement,   where    a    variety    of    fixation 
objects    has    been    used,    the    cross    is    found    to    give    the    best    control    of    fixation. 

I  I 

(2)        ■         or     ■    .    In  these  figures  also  the  test  is  to  distinguish  the  dot  clearly.    The 

II 
former  figure,   however,  is  a  little  too  complicated.     There  is  both  a  tendency  to  lose  the 
dot  and  for  the  lines  to  run  together  on  either  side.     A  simpler  criterion  gives  an  easier 
and  safer  judgment.     Doubtless  with  a  little  effort  other  figures  can  be  found  possessing 
still  greater  merit  as  test  objects. 


FERREE  :      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  45 

vision  is  determined  by  the  distance  at  which  he  can  just  clearly 
distinguish  in  every  detail  the  two  test  objects.  The  results  are 
thus  rendered  directly  in  terms  of  clearness  of  vision,  and  there 
are  no  progressive  errors  introduced  by  memory  and  fatigue. 

We  have  good  reason  to  believe  that  the  brightness  sensitivity, 
color  sensitivity,  and  visual  acuity  tests  rendered  sensitive  and 
adapted  to  our  purpose  in  the  manner  described  above  will  serve 
as  a  measure  of  the  general  level  of  efficiency  of  the  unfatigued 
eye  under  different  systems  of  illumination.  For  example,  they 
show  considerable  difference  in  result  when  the  tests  are  given 
under  three  types  of  lighting  now  in  use:  namely,  systems  of 
direct  lighting,  systems  of  indirect  lighting,  and  daylight.  In  each 
of  these  cases,  the  intensity  of  the  light  falling  on  the  test  object 
measured  in  foot-candles  is  kept  the  same.  The  tests  can  not, 
however,  be  depended  upon  to  show  a  loss  of  efficiency  of  the 
eye  as  a  result  of  three  or  four  hours  of  work  even  under  a  very 
unfavorable  lighting  system. 

III.  LOSS  OF  EFFICIENCY  AS   THE  RESULT  OF  A 
PERIOD  OF  WORK. 

We  have  no  reason  to  believe  that  the  brightness  and  color 
sensitivity  tests  have  failed  to  show  that  the  eye  loses  in  efficiency 
as  the  result  of  a  period  of  work  under  an  unfavorable  lighting 
system  because  of  any  fault  in  the  tests.  The  tests  used  are  the 
product  of  several  years  of  study  by  the  writer  of  the  sensitivity 
of  the  eye  to  brightness  and  color  and  of  the  factors  that  in- 
fluence this  sensitivity.  There  is  doubtless  very  little,  if  any,  loss 
of  sensitivity  during  this  length  of  time.  In  fact  it  is  commonly 
believed  that  the  brightness  and  color  processes  are  compensating 
in  nature.  The  case  is  quite  different,  however,  with  the  conven- 
tional visual  acuity  test,  or  even  with  the  modification  of  it  de- 
scribed above.  Although  brightness  and  color  sensitivity  are  fac- 
tors influencing  the  visual  acuity  test,  still  in  every  case  to  which 
it  may  be  applied,  it  is  predominantly  a  test  of  the  refracting 
mechanism  of  the  eye  and  its  muscular  control.  In  fact  our  re- 
sults for  the  tests  of  brightness  and  color  sensitivity  teach  us  that 
when  applied  to  the  case  in  hand  in  which  there  has  been  no 
change  in  the  quality  and  intensity  of  the  illumination  or  of  the 
refracting  mechanism  from  the  beginning  to  the  close  of  work, 


46  TRANSACTIONS  I.  E.  S. — PART  II 

the  results  of  the  visual  acuity  test  may  be  ascribed  practically 
entirely  to  changes  in  the  muscular  control  of  the  refracting 
mechanism,  or  at  least  to  changes  in  the  muscular  control  of  the 
eyes  as  a  whole. s  Now  the  visual  acuity  test,  when  it  is  confined 
to  a  momentary  judgment  of  clearness  of  vision,  is  not  adapted 
to  show  a  loss  in  muscular  efficiency  because,  although  this 
efficiency  may  have  been  lowered  enormously,  it  may  rise  momen- 
tarily under  the  spur  of  the  test  to  its  usual  level,  or  at  least  to 
the  level  obtaining  at  the  beginning  of  work.  Just  as  the  runner 
may,  under  the  spur  of  his  will,  equal  in  the  last  lap  of  his  course 
the  highest  speed  he  has  attained  at  any  other  point  in  the  course ; 
so  may  the  flagging  muscles  of  the  eye  be  whipped  up  to  their 
normal  power  long  enough  to  make  the  judgment  required  by  the 
visual  acuity  test.  It  was  the  feeling  of  all  our  observers  that  at 
the  close  of  work  under  the  system  of  direct  lighting  installed  in 
our  laboratory  the  eye  had  lost  heavily  in  efficiency.  A  great  deal 
of  discomfort  was  felt.  The  test  was  painful  and  was  accom- 
plished only  with  decided  strain.  Still  the  judgment  could  be 
made  apparently  with  as  much  accuracy  as  at  the  beginning  of 
work.  But  just  as  the  runner  finishing  his  course  cannot  long  keep 
up  his  extra  burst  of  speed,  so  might  we  expect  that  the  eye 
cannot  sustain  its  extra  effort.  This  analogy  led  the  writer 
to  continue  the  visual  acuity  test  through  an  interval  of  time. 
After  considerable  experimentation  an  interval  of  three  minutes 
was  chosen  as  best  suited  for  our  purpose.  Our  surmise  proved 
to  be  correct.  The  fatigued  eye  cannot  keep  up  its  extra  effort. 
The  results  of  the  test  showed  an  enormous  loss  of  efficiency  as 

8  Before  the  writer  would  speak  with  full  certaint3',  however,  that  the  retina 
loses  none  of  its  power  to  function  for  color  and  brightness  sensation  during  the  above 
stated  period  of  work,  he  would  feel  it  necessary  to  perform  another  kind  of  test  for  color 
and  brightness  sensitivity.  This  test  has  been  devised  by  him  especially  to  meet  the 
needs  of  this  problem.  In  this  test  the  element  of  time  is  introduced.  It  is  possible 
that  the  retina  may  have  lost  in  power  to  give  color  and  brightness  sensation  as  the  result 
of  a  period  of  work  even  when  the  conventional  test  based  on  a  momentary  judgment, 
shows  no  loss  of  sensitivity.  That  is,  it  may  be  more  susceptible  to  fatigue  as  the  result 
of  the  preceding  work.  To  determine  this,  a  fatigue  test  should  be  run  at  the  beginning 
and  close  of  work.  For  color  this  may  be  done  in  two  ways,  (i)  A  given  amount  of 
colored  light  may  be  used  and  the  time  required  for  the  eye  to  become  completely  ex- 
hausted or  insensitive  to  this  color  may  be  determined.  The  difference  in  time  required 
for  this  amount  of  fatigue  to  take  place  at  the  beginning  and  at  the  close  of  work  will 
represent  how  much  the  retina  has  lost  in  its  power  to  function  forcolor.  (2)  Theexperi- 
meni  need  not  be  continued  until  complete  exhaustion  takes  place.  The  amount  of 
exhaustion  that  has  taken  place  in  a  given  interval  of  time  can  be  measured.  As  before^ 
this  can  be  done  at  the  beginning  and  at  the  close  of  work  and  the  results  can  be  compared 
to  find  out  how  much  the  retina  has  lost  in  power  to  give  color  sensation. 


ferrEE:     tests  for  the  efficiency  of  the  eye  47 

the  consequence  of  three  hours  of  work  under  the  system  of 
direct  lighting,  while  in  daylight  practically  no  loss  was  shown. 

In  detail  the  test  is  as  follows.  When  the  observer  is  required 
to  look  at  the  test  card  for  three  minutes,  the  test  objects,  even 
when  the  eyes  are  fresh,  are  not  seen  clearly  for  the  whole  time. 
The  muscular  effort  required  to  keep  the  eyes  adjusted  for  clear 
vision  cannot  be  sustained  steadily  for  that  length  of  time.  The 
test  objects  are  seen  alternately  as  clear  and  blurred.  The  time 
they  are  seen  clear  and  blurred  is  recorded  on  a  rotating  drum 
upon  which  a  line  registering  seconds  is  also  run.  From  this 
record  the  ratio  of  the  time  seen  clear  to  the  time  seen  blurred 
is  determined.  This  ratio  may  be  fairly  taken  as  a  measure  of 
the  efficiency  of  the  eye  at  the  time  the  test  is  taken.  In  applying 
the  test  to  our  problem  the  record  is  taken  at  the  beginning  and 
at  the  close  of  work,  and  the  ratios  of  the  time  clear  and  the  time 
blurred  are  compared  for  the  two  cases  to  determine  how  much 
the  eye  has  lost  in  efficiency  as  a  result  of  work.  Two  values 
were  chosen  for  the  distance  at  which  the  test  card  was  placed 
from  the  eye:  (a)  the  maximal  distance  at  which  the  test  objects 
could  be  seen  clearly  in  the  momentary  judgment,  and  (b)  a 
distance  less  than  this.  The  latter  distance  was  chosen  because 
for  the  maximal  distance  towards  the  close  of  the  test,  even  when 
the  eyes  were  fresh,  the  value  of  the  time  blurred  became,  it  was 
thought,  excessively  high.  Results  for  the  two  distances,  there- 
fore, give  probably  a  fairer  expression  of  the  loss  in  efficiency 
than  for  the  one. 

The  problem  dealing  with  loss  of  efficiency  as  the  writer  has 
conceived  it  presents  two  phases.  We  may  investigate  (a) 
whether  the  eye  shows  a  loss  of  efficiency  after  three  or  four 
hours  of  work  under  a  given  lighting  system,  and  (b)  whether 
there  is  a  progressive  loss  of  efficiency  in  working  several 
months  or  years  under  a  given  lighting  system.  Only  the  first 
part  of  this  investigation  has  been  attempted  thus  far  in  our 
work  and  it  has  been  undertaken,  not  so  much  for  the  purpose 
of  making  an  exhaustive  study  of  loss  of  efficiency  under 
a  given  set  of  conditions,  as  it  has  been  to  get  a  sensitive  and 
practical  method  of  detecting  loss  of  efficiency.  In  order  to 
determine  whether  the  method  we  have  described  is  practical  and 
sufficiently  sensitive  for  our  purpose,  tests  should  be  made  on  a 
5 


48  TRANSACTIONS  I.  E.  S. — PART  II 

large  number  of  people  under  a  wide  range  of  lighting  condi- 
tions. We  have  not  as  yet  made  tests  under  a  wide  range  of 
lighting  conditions.  We  have  chosen  rather  to  begin  with  three 
broad  types  of  illumination  now  in  general  use ;  systems  of  direct 
lighting,  systems  of  indirect  lighting,  and  daylight.  Types  based 
upon  the  distribution  of  light  have  been  selected  because  it  has 
seemed  to  the  writer,  both  from  his  own  work  and  from  a  survey 
of  the  work  done  by  others,  that  distribution  or  diffuseness  of 
light  is  the  most  important  factor  we  have  yet  to  deal  with  in 
our  search  for  conditions  that  give  minimum  loss  of  efficiency 
and  maximum  comfort  in  seeing.  The  quality  of  the  light  and 
its  intensity  at  the  source  are  already  pretty  well  taken  care  of, 
apparently  at  least  better  taken  care  of  in  general  practise,  rela- 
tive to  their  importance  to  the  eye,  than  is  distribution.  A 
detailed  report  of  our  results  will  not  be  given  in  this  paper. 
The  following  results  selected  as  typical  from  a  large  number 
of  observations  are  appended,  however,  to  show  how  the  effi- 
ciency of  the  eye  as  measured  by  the  above  test  falls  off  as  the 
result  of  three  hours  of  work  under  a  system  of  direct  lighting 
as  compared  with  daylight. 

The  tests  were  conducted  in  a  room  30.5  feet  (9.29  m.)  long, 
22.3  feet  (6.797  m0  wide  and  9.5  feet  (2.895  m-)  high.  The 
daylight  illumination  came  from  six  windows,  all  on  one  side 
provided  with  thin  white  curtains  to  secure  the  necessary  control. 
The  artificial  lighting9  was  accomplished  by  means  of  two  rows 
of  fixtures  of  four  fixtures  each.  Each  row  was  6  feet  (1.828 
m.)  from  the  side  wall  and  the  fixtures  were  6  feet  apart.  Each 
fixture  was  supplied  with  two  16  candle-power  carbon  lamps 
29  inches  (0.736  m.)  from  the  ceiling  with  a  white  porcelain 
reflector  16  inches  (0.406  m.)  in  diameter  fastened  directly  above. 
The  daylight  tests  were  made  at  9  a.  m.  and  12  m.  Between  these 
limiting  times,  the  observer  was  required  to  read  pages  of  type, 
uniform  in  size,  printed  upon  paper  of  uniform  texture  of  sur- 
face and  of  uniform  reflecting  power.    The  tests  for  the  system 


9  This  room  gave  the  impression  of  being  brilliantly  lighted.  The  writer  was  amazed 
to  find,  however,  that  only  2.5  foot-candles  of  light  were  received  on  the  test  card  placed 
about  midway  between  two  of  the  rows  of  lights  and  midwa5'  between  two  sets  of  fixtures. 
The  walls  and  ceiling  of  the  room  were  of  plaster,  natural  finish,  and  the  floor  of  dark 
tiling.  Before  our  tests  were  taken,  the  walls  and  ceilings  were  painted  white  which 
nearly  doubled  the  light  received  on  the  test  card. 


FERREE  :      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  49 

of  direct  lighting  were  taken  at  7  p.  m.  and  10  p.  m.  During  the 
interval  intervening,  the  observer  was  required  to  read  type  of 
the  same  size  and  printed  on  the  same  paper  as  was  used  in  the 
daylight  work.  The  reading  was  done  in  each  case  at  exactly 
the  same  spot  in  the  room  as  at  which  the  tests  were  made.  The 
intensity  of  illumination  was  also  in  both  cases  made  as  nearly 
equal  as  it  was  possible  to  do  by  methods  now  available.10  The 
two  tests  were  always  given  on  successive  days  but  one.  In  order 
to  guarantee  that  the  observer's  physical  and  optical  condition 
should  be  as  nearly  the  same  for  the  two  tests  as  it  was  possible 
to  obtain,  he  was  required  to  rest  during  the  day  immediately 
preceding  each  test.  Since  the  li  test  has  proven  to  be  the  more 
sensitive,  results  will  be  given  for  it  alone  in  the  following  table. 
Column  1  of  this  table  gives  the  time  of  day  at  which  the  work 
was  done  and  the  tests  were  made.  Column  2  gives  the  type  of 
test.  Column  3  gives  the  distance  of  the  test  card  from  the  eye. 
As  stated  earlier  in  the  paper,  two  distances  were  used;  one  the 
maximum  at  which  the  test  object  could  be  seen  clearly,  the  other 
a  distance  less  than  this.  Division  A  of  the  table  gives  the  results 
for  the  former  distance ;  division  B,  for  the  latter.  Columns  4 
and  5  respectively,  give  the  number  of  times  the  test  object  was 
seen  clear  and  unclear.  Column  6  gives  the  number  of  seconds 
in  the  three  minutes  that  the  test  object  was  seen  clear,  and 


10  In  order  to  equalize  the  intensity  of  illumination,  a  method  of  measurement  is 
required.  Two  methods  were  used  by  us  ;  photometry,  and  a  more  delicate  method 
based  upon  the  sensitivity  of  the  peripheral  retina  to  brightness  contrast.  In  case  of  the 
former,  a  Sharp-Millar  portable  photometer  was  used.  The  light  falling  upon  the  test 
card  was  measured  in  foot-candles  and  was  made  equal  for  each  type  of  lighting.  Full 
details  of  the  latter  method  will  not  be  given  here.  As  stated  above  it  is  based  upon  the 
extreme  sensitivity  of  the  peripheral  retina  to  brightness  contrast,  especially  to  the 
induction  by  a  white  screen.  To  apply  the  method,  some  given  illumination  is  taken 
as  standard.  The  amount  of  black  induced  by  a  white  campimeter  screen  upon  a 
15  mm.  area  of  some  medium  gray,  {e.  g.  Hering  gray  No.  14)  at  an  excentricity  of  25 
deg.  in  the  temporal  meridian,  is  measured.  This  amount  of  contrast  is  taken  as  the 
index  of  that  illumination.  To  duplicate  the  illumination  at  any  succeeding  time,  the 
intensity  is  varied  until  the  same  amount  of  contrast  is  induced  by  the  white  screen  on 
the  gray  at  the  25  deg.  point,  for  the  same  observer.  This  method  was  devised  in  the 
writer's  laboratory  and  he  has  found  by  repeated  trials  that,  although  it  is  not  so  con- 
venient for  many  of  the  purposes  for  which  the  photometric  method  is  used,  it  is  many 
times  more  sensitive  than  the  traditional  photometric  method.  The  Sharp-Millar 
photometer,  like  other  photometers,  is  insensitive  for  the  determination  of  the  illumina- 
tion of  a  room  by  daylight.  This  is  because  the  standard  field  illuminated  by  the 
tungsten  lamp  is  deep  orange  in  color,  while  the  comparison  field  illuminated  by  day- 
light is  clear  white.  This  difference  in  color  tone  makes  the  judgment  of  brightness 
equality  difficult  to  make  and  renders  the  instrument  extremely  insensitive  for  daylight 
work. 


50  TRANSACTIONS  I.  E.  S. — PART  II 

column  7  the  number  of  seconds  unclear.  Column  8  gives  the 
ratio  of  the  total  time  clear  to  the  total  time  unclear.  This  ratio 
as  stated  earlier  in  the  paper  expresses  the  efficiency  of  the  eye 
for  clear  seeing  for  an  interval  of  three  minutes  at  the  time  at 
which  the  test  was  taken. 

TABLE  I. 
Showing  How  the  Eye  Falls  Off  in  Efficiency  as  the  Result  of  Three  Hours 

of  Work  under  a  System  of  Direct  Lighting  as  Compared  with  Daylight. 

In  Division  A  the  Test  Card  is  Put  at  the  Maximal  Distance  at  Which  the 

Test  Object  Could  be  Seen  Clearly;  in  Division  B,  at  a  Distance  Less  than 

This.11 


Time 

of 

day 

Test 

Distance 

of  card 

from  eye 

cm. 

Number 

of 
times 
clear 

Number 

of 

times 

unclear 

Total 

time 

clear 

sec. 

Total 

time 

unclear 

sec. 

Total  time  clear 
Total  time  unclear 

A. 

9  A.  M. 

li 

102 

15 

15 

105.6 

78.4 

1.4 

12  M. 

li 

102 

15 

14 

103. 1 

76.9 

i-33 

7  P.  M. 

li 

75 

18 

18 

II9.7 

60.3 

1.98 

IO  P.  M. 

B. 

9  A.  M. 

li 

75 

15 

15 

55-4 

124.6 

0.44 

li 

92 

14 

13 

136.8 

43-2 

3.16 

12  M. 

li 

92 

12 

12 

134-9 

45-i 

2.99 

7  P.  M. 

li 

65 

24 

23 

141.8 

38.2 

3-7 

IO  P.  M. 

li 

65 

17 

17 

75-5 

104.5 

0.72 

11  It  will  be  noticed  in  the  table  that  the  ratio  total  time  seen  clear  -f-  total  time  seen 
unclear  is  smaller  for  the  test  both  at  the  beginning  and  at  the  close  of  work  in  division 
A  where  the  maximal  distance  at  which  the  test  object  could  be  seen  was  used,  than  in 
division  B  where  a  distance  less  than  this  was  used.  This  is  just  what  should  be  expected 
from  the  nature  of  the  test.  For  it  may  be  said  that,  within  limits,  the  nearer  the  object 
is  to  the  eye  the  greater  is  the  proportion  of  time  it  should  be  seen  clearly  ;  and,  con- 
versely, the  farther  the  object  is  from  the  eye  the  smaller  is  the  proportion  of  time  it 
should  be  seen  clearly.  It  will  also  be  noticed  that  the  ratio  is  slightly  larger  when  the 
tests  are  made  under  the  system  of  direct  lighting  than  when  made  under  daylight.  The 
explanation  of  this,  too,  is  found  in  terms  of  the  distances  that  were  chosen  for  the  test 
object.  These  distances,  relative  to  the  maximal  distance,  were  chosen  shorter  for  the 
artificial  light  than  for  daylight.  This  was  done  because  of  the  large  falling  off  in  the 
ratio  gotten  for  the  test  at  the  close  of  work  under  the  artificial  light.  Had  the  first  ofthe 
two  distances  used  in  these  tests,  for  example,  been  chosen  as  near  to  the  maximal  dis- 
tance for  the  artificial  light  as  it  was  for  daylight,  the  result  ofthe  test  made  at  the  close 
of  work  would  have  been  that,  after  the  first  interval  seen  as  clear,  the  test  object  would 
have  been  seen  unclearly  during  the  remainder  of  the  test.  At  first  glance  one  might 
be  tempted  to  think  that  the  difference  in  the  scale  of  magnitude  for  the  two  ratios,  is  due 
to  some  inequality  in  the  intensity  of  the  illumination  that  was  given  by  the  two  systems 
of  lighting.  It  is  obvious  on  reflection,  however,  that  the  intensity  of  illumination  can 
have  little  or  nothing  to  do  with  the  scale  of  magnitude  of  these  ratios.  The  intensity  of 
the  illumination  influences  the  maximal  distances  at  which  the  test  object  can  be  seen 
clearly  but  the  scale  of  magnitude  of  the  ratio,  time  clear  to  time  unclear  must  depend 
primarily  upon  how  near  the  distance  chosen  for  the  test  object  is  to  the  maximal 
distance.  (This  principle,  it  is  obvious,  does  not  affect  the  comparison  of  the  ratios 
obtained  at  the  beginning  and  close  of  work  under  a  given  lighting  system  for  the  dis- 
( Continued  on  following  page. ) 


FERREE:      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE 


51 


In  order  to  give  a  typical  representation  in  graphic  form  of 
the  effect  of  three  hours  of  work  on  the  efficiency  of  the  eye  in 
daylight  and  under  the  system  of  direct  lighting,  estimated  in 
terms  of  the  test  we  have  described,  the  results  of  the  above  table 
are  given  in  the  form  of  a  curve.  In  constructing  this  curve  the 
length  of  time  of  work  is  plotted  along  the  abscissa  and  ratio  of 
the  time  the  test  object  is  seen  clear  to  the  time  unclear,  is  plotted 
along  the  ordinate.  Each  one  of  the  large  squares  along  the 
abscissa  represents  an  hour  of  work,  and  along  the  ordinate  an 
integer  of  the  ratio.  Figure  I  shows  the  result  of  division  A  and 
figure  II  for  division  B  of  the  table.     An  inspection  of  these 


\z.m 

M. 

mooRM. 

i 

7.00  RK 
aooA.M. 


Fig.  I.— Curve  for  division  A  of  the  table.    Showing  how  the  eye  falls  off  in  efficiency 
for  three  hours  of  work  under  a  system  of  direct  lighting  as  compared  with  daylight. 

curves  shows  that  the  efficiency  of  the  eye  measured  by  the  ratio 
of  the  time  the  test  objects  are  seen  clear  to  the  time  seen  unclear, 
falls  off  rapidly  for  the  system  of  direct  lighting  but  scarcely  at 
all  for  daylight. 

Although  it  has  been  the  purpose  of  this  paper  merely  to  out- 


tance,  once  it  is  chosen  for  that  system,  is  kept  the  same  or  both  tests).  As  further 
proof  that  the  difference  in  the  intensity  of  illumination  had  nothing  whatever  to  do 
with  this  result,  the  intensity  of  illumination  was  carefully  determined  immediately 
before  and  after  these  tests  and,  if  the  readings  showed  any  inconstancy  in  the  illumi- 
nation, the  results  were  discarded  and  new  tests  were  made.  The  above  explanation 
should  be  borne  in  mind  also  in  examining  the  curves  plotted  from  the  results  of  the  table. 
The  curve  for  division  B  of  the  table,  for  example,  begins  at  a  higher  point  on  the  ordi- 
nate than  for  division  A  ;  and  the  curve  for  the  artificial  illumination  starts  at  a  higher 
point  than  the  curve  for  daylight. 

It  is  scarcely  necessary  to  point  out  that  neither  the  scale  of  magnitude  of  ratio  nor 
the  point  at  which  the  curve  starts  is  of  any  considerable  consequence  for  our  work.  The 
important  thing  is  not  how  large  is  the  ratio  at  the  beginning  of  work,  but  how  much  it 
falls  off  as  the  result  of  work.  In  fact  the  magnitude  of  ratio  need  not  be  taken  into  ac- 
count at  all  any  further  than  that  it  chances  to  be  a  coincident  result  of  a  condition  that 
seems  to  render  our  test  more  sensitive.  That  is,  our  results  seem  to  show  that  the  ratio 
falls  off  more  when  the  distance  chosen  for  the  test  object  is  not  too  near  the  maximal  dis- 
tance. In  future  work,  therefore,  more  care  should  be  taken  probably  than  was  exercised 
in  this  preliminary  study  to  choose  the  distances  for  the  test  object  so  that  in  case  of  each 
lighting  system  employed  they  shall  sustain  the  same  ratio  to  their  corresponding 
maximal  distances. 


52 


TRANSACTIONS  I.  E.  S. — PART  II 


line  and  in  part  to  demonstrate  a  set  of  tests,  a  word  of  dis- 
cussion and  interpretation  of  the  results  we  have  reported  may 
not  be  out  of  place  here.  Since  the  visual  acuity  test  (given 
under  constant,  quality,  intensity  and  distribution  of  light)  is  a 
test  largely  of  the  refractive  mechanism  of  the  eye  and  its  mus- 
cular control  and  since  the  refractive  mechanism  could  not  have 
changed  during  three  or  four  hours  of  work,  the  obvious  indi- 
cation of  the  above  result  is  that  the  loss  of  efficiency  sustained 
by  the  eye  in  these  experiments  is  a  loss  in  muscular  efficiency. 
This  conclusion  is  borne  out  also  by  the  fact,  stated  earlier  in 
the  paper,  that  the  direct  tests  of  the  efficiency  of  the  retina, 
namely,  the  test  for  brightness  and  color  sensitivity  did  not  show 
conclusively  any  loss.12  Moreover,  the  conlusion  is  in  line  with 
current  conception.  In  current  theory  the  retina  is  considered 
as  a  mechanism  more  or  less  compensating  in  its  action,  while 


700  RM. 

\ 

9.00  A.M. 

\ 

s 

I2.0C 

)M. 

IOloo 

RM. 

Fig.  II.— Curve  for  division  B  of  the  table.  Showing  how  the  eye  falls  off  in  efficiency 
as  the  result  of  three  hours  of  work  under  a  system  of  direct  lighting  as  compared 
with  daylight. 

the  muscles  of  the  eye  are  not  so  considered.  The  following 
reasons  are  suggested  why  the  muscles  of  the  eye  giving  both 
fixation  and  accommodation  should  be  subjected  to  a  greater 
strain  by  the  system  of  direct  lighting  than  by  daylight.  ( i )  The 
bright  images  of  the  electric  bulbs  falling  on  the  peripheral  retina 
which  is  in  a  perpetual  state  of  darkness  adaptation  as  compared 
with  the  central  retina  and  is  therefore  extremely  sensitive  in  its 
reaction  to  such  intensive  stimuli,  sets  up  a  reflex  tendency  for 
the  eye  to  fixate  them  instead  of  the  letters  which  the  observer 
is  engaged  in  reading.  (2)  Likewise,  a  strong  reflex  tendency 
to  accommodate  for  these  brilliant  sources  of  light  all  at  differ- 


13  This  statement  is  also  subject  to  the  foot-note  appended  to  the  earlier  statement. 


FERREE  :      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  53 

ent  distances  from  each  other  and  from  the  lettered  page,  is  set 
UP-  (3)  These  brilliant  images  falling  upon  a  part  of  the  retina 
that  is  not  adapted  to  them  causing  as  they  do  acute  discomfort 
in  a  very  short  period  of  time,13  doubtless  induce  spasmodic  con- 
tractions of  the  muscles  which  both  disturb  the  clearness  of 
vision  and  greatly  accentuate  the  fatiguing  of  the  muscles.  The 
net  result  of  all  these  causes  is  excessive  muscular  strain  which 
soon  shows  itself  as  a  loss  in  power  to  do  work.  In  the  illumina- 
tion of  a  room  by  daylight  with  a  proper  distribution  of  windows, 
the  situation  is  quite  different.  The  field  of  vision  contains  no 
bright  sources  of  light  to  distract  fixation  and  accommodation 
and  to  cause  spasmodic  muscular  disturbances,  due  to  the  action 
of  intensive  light  sources  upon  the  dark  adapted  and  sensitive 
peripheral  retina.  In  daylight  the  light  waves  have  suffered 
innumerable  reflections  and  the  light  has  become  diffuse.  The 
field  of  vision  is  uniformly  illuminated.  The  illumination  of  the 
retina,  therefore,  falls  off  more  or  less  uniformly  from  fovea  to 
periphery  as  it  should  in  order  to  permit  of  fixation  and  accom- 
modation for  a  given  object  with  the  minimum  amount  of  strain. 

It  is  not  our  purpose  to  contend  in  this  report  that  distribution 
is  the  only  factor  of  importance  in  the  illumination  of  a  room. 
The  intensity  and  the  quality  of  the  illumination  must  also  be 
taken  into  consideration.  To  test  the  relative  effect  of  these 
factors  upon  the  working  power  of  the  eye,  records  would  have 
to  be  taken  when  each  was  varied  in  turn  and  the  other  two 
maintained  constant.  In  the  results  shown  in  the  above  tables 
the  intensity  alone  was  constant  in  the  two  cases.  Both  the 
quality  and  the  distribution  were  different  in  the  direct  lighting 
system  and  the  illumination  by  daylight.  The  difference  in  the 
results  obtained  will  have,  therefore,  to  be  attributed  both  to 


13  There  is  no  doubt  in  the  writer's  mind  that  the  eye-discomfort  experienced  as  the 
result  of  work  under  an  unfavorable  system  of  lighting  is  not  by  any  means  all  muscular. 
The  "sandiness"  passing  over  into  a  stinging,  stabbing  pain  which  comes  early,  in  the 
experience  of  discomfort  seems  to  be  conjunctival.  And  while  the  retina  itself  is 
apparently  insensitive  to  pain  from  mechanical  stimulation,  still  when  exposed  to  a 
source  of  light  of  a  brilliancy  to  which  it  is  not  adapted," a  painful  reaction  is  produced 
which  can  scarcely  be  considered  muscular.  For  example,  after  confinement  for  some 
time  in  a  dark-room  exposure  to  ordinary  daylight  is  painful  to  the  normal  e3-e.  That 
this  is  not  entirely  muscular  can  be  shown  by  the  fact  that  a  similar  reaction  is  experi- 
enced when  the  ciliary  and  iris  muscles  are  paralyzed  by  atropine.  The  reaction  is  also 
experienced  by  aphakial  subjects  whose  lenses  have  been  so  long  removed  that  muscular 
atrophy  must  have  taken  place. 


54  TRANSACTIONS  I.  E.  S. — PART  II 

difference  in  the  distribution  and  to  difference  in  the  quality  of 
the  illumination.  In  our  tests  comparative  of  the  systems  of 
direct  and  indirect  lighting,  the  results  of  which  will  be  reported 
in  a  later  paper,  clear  tungsten  lamps  will  be  used  in  both  cases. 
The  intensity  will  be  made  the  same  and  the  quality  of  the  light 
will  be  approximately  the  same.  The  distribution  or  diffusion 
alone  will  be  different.  Whatever  difference  in  result  we  get  in 
these  two  cases  can,  therefore,  with  reasonable  certainty  be 
attributed  to  the  differences  in  the  distribution  of  light. 

With  regard  to  the  effect  of  varying  the  intensity  of  illumina- 
tion, our  results  show  nothing;  with  regard  to  the  effect  of  varying 
quality,  nothing  in  isolation;  and  with  regard  to  distribution,  we 
have  data  only  for  such  differences  as  are  found  in  the  three 
types  of  illumination  now  in  general  use.  In  later  work,  however, 
the  analysis  along  these  lines  will  be  completed.  We  hope  on 
the  laboratory  side,  to  make  a  systematic  study  of  the  effect  of 
wide  ranges  of  variation  of  each  of  the  factors  in  turn.  It  will 
be  comparatively  easy,  for  example,  to  keep  the  intensity  and 
distribution  constant  and  vary  the  quality,  or  to  keep  the  quality 
and  distribution  constant  and  vary  the  intensity.  We  hope  in 
addition,  to  supplement  this  work  by  testing  the  eyes  of  employ- 
ees who  work  under  a.  given  lighting  system  for  several  hours  a 
day,  for  evidences  of  a  progressive  loss  of  efficiency. 

IV.  A  PRELIMINARY  STUDY  OF  THE  CAUSES  OF 
DISCOMFORT. 

In  addition  to  studying  the  conditions  that  give  us  maximal 
efficiency,  it  is  important  to  determine  the  lighting  conditions 
and  eye  factors  that  cause  discomfort.  In  fact,  it  might  well  be 
said  that  our  problem  in  lighting  at  present  is  not  so  much  how 
to  see  better  as  it  is  how  to  see  with  more  comfort  and  with  less 
damage  to  the  general  health  on  account  of  eye  strain.  Any 
comparative  study  of  the  conditions  producing  discomfort  neces- 
sitates a  means  of  estimating  discomfort.  It  is  obvious  that  the 
core  of  the  experience  of  discomfort  is  either  a  sensation  or  a 
complex  of  sensations.  As  such,  it  should  have  a  limen  or 
threshold  value  just  as  other  sensations  have;  and  just  as  we  are 
able  in  general  to  estimate  sensitivity  in  terms  of  the  threshold 
value,  so  should  we  in  this  case  be  able  to  use  the  threshold 


FERREE  :      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  55 

value  in  estimating  the  eye's  sensitivity  or  liability  to  discomfort 
under  a  given  lighting  condition.  Threshold  values  are  usually 
determined  by  finding  how  much  energy  or  intensity  of  a  given 
stimulus  applied  for  a  short  interval  of  time  is  required  to  arouse 
a  just  noticeable  sensation.  This  form  of  procedure,  however, 
is  not  adapted  to  the  needs  of  our  problem.  It  is  much  better  to 
reverse  the  process  and  find  how  long  the  eye  has  to  be  exposed 
to  a  stimulus  of  a  given  intensity  to  arouse  just  noticeable  dis- 
comfort. Our  limen,  then,  becomes  a  time  limen,  and  is  meas- 
ured in  units  of  time  instead  of  in  units  of  intensity.  In  order 
to  determine  whether  the  judgment  of  the  limen  of  discomfort 
can  be  made  with  certainty  and  to  test  in  general  the  feasibility 
of  the  method,  the  writer  undertook  to  determine  the  compara- 
tive sensitivity  of  the  eye  to  discomfort  when  the  source  of  light 
was  exposed  in  different  parts  of  the  field  of  vision.  In  order  to 
carry  out  this  investigation,  a  16  candle-power  lamp  was  attached 
to  the  arm  of  a  perimeter  in  such  a  way  that  the  end  of  the  bulb 
was  always  directed  towards  the  observer's  eye.  The  arm  of  the 
perimeter  could  be  shifted  to  any  meridian  in  which  it  was  desired 
to  work,  and  the  lamp  could  be  moved  at  will  along  the  arm.  It 
was  thus  possible  to  expose  the  light  at  any  point  in  the  field  of 
vision  that  was  desired.  Working  in  this  way,  we  have  investi- 
gated the  effect  of  many  types  of  variation  of  the  distribution  of 
the  light  in  the  visual  field,  and  it  is  our  purpose  to  extend  the 
investigation  as  fast  as  possible  to  the  variation  of  the  other 
factors.  Of  the  variations  we  have  made  in  the  distribution  of 
the  light  in  the  field  of  vision,  it  will  be  necessary,  however,  in 
order  to  illustrate  the  general  method  of  working,  to  describe 
only  one,  namely,  the  exposure  of  the  source  of  light  at  different 
points  in  the  field  of  vision  for  one  eye  when  fixation  and  accom- 
modation were  taken  for  a  far  point. 

In  carrying  out  the  investigation,  the  following  precautions 
were  observed,  (a)  It  was  found  better  to  work  in  a  room  mod- 
erately illuminated  by  a  source  of  light  behind  the  observer  and 
entirely  concealed  from  him  rather  than  in  the  dark.  The  inter- 
vals of  dark-adaptation  between  exposures  in  the  dark-room 
seemed  to  make  the  observer's  eye  too  sensitive  for  our  purpose. 
This  was  especially  true  for  certain  parts  of  the  peripheral  retina. 
6 


56  TRANSACTIONS  I.  E.  S. — PART  II 

In  becoming  supersensitive  there  was  a  tendency  to  become 
erratically  sensitive,  (b)  It  was  found  that  blinking  serves  as  a 
variable  factor  for  the  relief  of  discomfort  and  that  the  amount 
of  blinking  must  be  made  constant  from  test  to  test.  This  was 
accomplished  by  having  the  observer  blink  at  equal  intervals  dur- 
ing the  exposure,  timing  himself  by  the  stroke  of  a  metronome. 
The  interval  most  natural  and  suitable  for  this  purpose  was  de- 
termined for  each  observer  separately,  (c)  All  comparisons  were 
planned  in  series.  For  example,  if  it  were  desired  to  compare 
the  sensitivity  of  the  temporal  and  nasal  halves  of  the  retina  in  a 
given  meridian,  the  exposure  was  first  made  at  a  given  point  in 
one  half  and  next  at  the  corresponding  point  in  the  other  half. 
This  was  to  guarantee  that  the  eye  should  be  as  nearly  in  the 
same  condition  with  regard  to  progressive  fatigue,  etc.,  as  was 
possible.  Further  to  safeguard  against  error  in  this  regard  ser- 
ies were  compared  in  which  the  exposures  were  repeated  in  the 
reverse  order,  (d)  An  interval  of  recovery  was  allowed  between 
exposures.  This  interval  had  to  be  determined  separately  for 
each  observer  and  often  had  to  be  made  different  for  the  same 
observer  on  different  days.  It  was  never  changed,  however, 
during  the  course  of  an  experiment,  the  results  of  which  were  to 
be  compared,  (e)  In  order  that  the  observer's  head  be  held 
rigidly  in  position  during  the  exposure,  he  was  required  to  bite 
an  impression  of  his  teeth  previously  made  and  hardened  in  wax 
on  a  mouthboard.  When  an  exposure  was  to  be  made,  the  fixa- 
tion was  taken,  the  light  turned  on,  and  a  signal  was  given  by  the 
observer  when  a  just  noticeable  discomfort  was  aroused,  or,  if 
it  was  desired,  when  the  different  stages  of  discomfort  were 
reached.  The  judgment  was  found  to  present  no  especial  diffi- 
culty, and  the  method,  when  properly  applied,  to  provide  a  feas- 
ible means  for  comparing  the  sensitivity  of  the  eye  to  discom- 
fort under  all  the  conditions  to  which  we  have  been  able  thus  far 
to  extend  its  application.  In  actual  practise  the  method  also 
brings  out  an  analysis  of  discomfort. 

Discomfort  seems  to  be  a  complex  of  three  experiences,  each  of 
which  develops  at  a  different  time.  When  the  light  is  turned  on,  we 
have  at  once  glare.  This  is  a  light  sensation  and  though  unpleas- 
ant has  no  painful  elements.   Next  comes  a  conjunctival  sensation 


FERREE  :      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  57 

which  begins  with  what  is  commonly  called  "sandiness"  and  soon 
passes  over  into  a  sharp,  stinging,  stabbing  pain.  Lastly  there 
comes  what  is  probably  a  muscular  discomfort, — a  hurting  and 
aching  in  the  ball  of  the  eye  which  if  the  exposure  is  continued 
long  enough  seems  to  radiate  to  the  socket  and  the  surrounding 
regions  of  the  face  and  head,  the  arch  of  the  brow,  the  forehead, 
the  temples,  etc.  Details  will  not  be  given  here  of  the  compara- 
tive sensitivity  of  different  points  of  the  retina  to  discomfort.  It 
will  be  sufficient  to  say,  that  the  periphery  of  the  retina  is  more 
sensitive  than  the  center;  that  the  nasal  half  is  in  general  more 
sensitive  than  the  temporal  half  and  the  upper  half  than  the 
lower  half ;  and  that  in  passing  from  the  center  to  the  periphery 
of  the  retina,  the  sensitivity  is  found  first  to  increase  then  to  de- 
crease, becoming  extremely  little  at  the  limits  of  the  field  of 
vision.  In  the  horizontal  meridians  both  on  the  temporal  and 
nasal  sides,  maximal  sensitivity  is  found  around  the  45  deg. 
point.  In  the  vertical  meridians,  maximal  sensitivity  seems  to 
be  near  the  point  15  deg.  below  the  horizontal.  In  a  paper  soon 
to  be  published,  a  detailed  statement  and  explanation  of  these 
results  will  be  given. 

DISCUSSION. 

Dr.  H.  E.  Ives:  This  paper  is  well  worth  the  while  of  pro- 
fessional psychologists  to  study ;  and  it  gives  to  the  illuminating 
engineer  results  which  are  extremely  valuable. 

When  the  illuminating  engineer  has  the  problem  of  producing 
satisfactory  results,  he  has  two  methods  of  doing  so ;  first,  the 
case  system,  in  which  he  copies  an  illumination  produced  by 
nature  or  invention,  which  has  proved  satisfactory  by  experience; 
and  he  hopes  to  get  the  same  result.  But  there  are  defects  in 
this  method ;  we  are  very  apt  to  follow  the  example  of  the  Chinese 
who  made  motors  by  copying  the  imported  ones  even  down  to 
the  color  of  the  paint  on  the  casings  and  the  scratches  on  the  paint 
We  may  do  equally  foolish  things  by  slavish  copying.  That  is  in- 
herent in  the  case  system.  Up  to  recently  some  of  us  have  been  of 
the  opinion  that  even  with  the  defects  of  this  case  system  we  could 
apply  it  to  advantage,  for  instance  by  studying  how  nature  pro- 
duces her  lighting  schemes.     But  in  order  to  make  any  great 


58  TRANSACTIONS  I.  t.  S. — PART  II 

progress  we  must  deviate  from  what  exists ;  we  must  experiment 
and  invent.  This  necessitates  some  means  of  testing  our  results 
and  this  process  of  experiment  and  test  constitute  the  second  pro- 
cedure. Our  Society  has  lately  been  interested  in  the  physiological 
side  of  illumination,  but  has  been  sadly  handicapped  by  the  lack  of 
significant  tests — we  have  been  dependent  practically  on  labor- 
iously acquired  experience.  One  great  object  in  adopting  a 
method  of  measurement  is  the  saving  of  time.  For  instance, 
suppose  our  only  means  of  measuring  voltage  was  by  the  duration 
of  physiological  disturbances  following  an  electric  shock.  In  order 
to  duplicate  a  voltage  which  gives  a  shock  whose  after  effects 
last  a  day,  we  would  require  weeks  or  months  of  toil,  because  of 
the  time  necessary  to  wait  for  the  results  of  each  experiment. 
Suppose  the  first  time  we  secured  the  desired  voltage  we  had  an 
instrument  known  as  a  "voltmeter;"  it  would  only  take  a  minute 
to  determine  that  voltage.  We  want  something  for  measuring 
the  effect  of  lighting  systems  which  will  enable  us  to  get  results 
with  a  speed  comparable  with  that  of  a  voltmeter. 

Various  methods  of  test  have  been  proposed  and  Dr.  Ferree 
has  gone  over  all  of  these.  He  arrives  at  a  conclusion  which  I 
think  it  behooves  us  all  to  observe;  namely,  that  these  tests  will 
show  what  he  calls  "the  general  level  or  scale  of  visual  efficiency," 
but  they  are  practically  useless  as  tests  of  the  loss  of  visual 
efficiency. 

Here  is  a  sentence  which  means  a  great  deal  "Just  as  a  runner 
may,  under  the  spur  of  his  will,  equal  in  the  last  lap  of  his  course 
the  highest  speed  he  has  attained  at  any  other  point  in  the  course, 
so  may  the  flagging  muscles  of  the  eye  be  whipped  up  to  their 
normal  power  long  enough  to  make  the  judgment  required  by  the 
visual  acuity  test." 

Dr.  Ferree  here  gives  us  the  benefit  of  his  point  of  view  and 
experience  in  these  matters.  In  this  paper  he  has  recognized  the 
inefficiency  of  the  methods  now  used.  He  realizes  that  we  want 
a  test  of  the  loss  of  visual  efficiency.  The  eye  may  respond 
momentarily,  like  the  tired  runner,  and  see  the  object  as  dis- 
tinctly as  before,  but  we  know  that  it  is  not  as  efficient.  Dr. 
Ferree  has  devised  a  test  in  which  is  introduced  a  time  element. 
The    observer   views    a   visual    acuity   test    object.     When    the 


FERREE  :      TESTS  FOR  THE  EFFICIENCY  OF  THE  EYE  59 

limit  of  visibility  is  found  the  observer  is  not  allowed  to  rest, 
because  he  will  again  after  an  interval  get  just  as  good  results 
as  at  first ;  instead  he  presses  a  key  as  long  as  the  detail  is  clear ; 
then  when  the  tired  muscles  flag  and  the  object  blurs,  the  finger 
on  the  key  is  removed.  At  first  it  appears  easy  to  see  the  detail 
clearly,  but  pretty  soon  it  is  not  so  easy  and  one  does  not  distin- 
guish the  chart  so  well.  Very  soon  it  becomes  necessary  to  take  the 
finger  off  the  key.  Intervals  of  clear  and  blurred  vision  alter- 
nate and  at  the  end  we  have  a  ratio  of  the  time  the  chart  is  dis- 
tinguishable to  the  time  when  it  is  not. 

Dr.  Ferree  has  tried  out  daylight  and  a  direct  artificial 
lighting  system  and  we  have  here  for  the  first  time  the  results  of 
that  test.  They  show  what  many  of  us  have  been  sure  of ; 
that  daylight  does  not  decrease  the  efficiency  nearly  as  much  as 
artificial  lighting.  On  the  fourteenth  and  fifteenth  pages  are  two 
charts  showing  by  straight  lines  the  falling  off  in  efficiency  which 
occurs  under  artificial  lighting  as  compared  to  daylight.  Person- 
ally, I  think  we  should  say  "Eureka!" 

I  hope  Dr.  Ferree  will  proceed  to  standardize  these  tests  and 
tell  us  the  best  working  distances  and  one  thing  and  another. 
As  he  is  not  here,  I  have  tried  to  bring  out  the  most  important 
points.  He  has  given  us  a  most  valuable  contribution,  and  I 
hope  before  long  we  will  be  in  a  position  to  settle  these  questions 
of  light  and  dark  walls  by  this  method  of  test  and  not  by 
"Kilkennycat"  discussion,  which  brings  us  nowhere. 

I  think  we  should  do  our  best  to  aid  Dr.  Ferree  to  develop  this 
method  of  test  to  give  us  what  now  we  can  get  only  by  experience. 
I  am  aware  that  I  have  not  done  this  paper  justice,  but  I  want  to 
express  my  appreciation  of  his  work. 

Mr.  C.  O.  Bond:  The  American  Medical  Association  is  a 
body  whose  conclusions  as  to  the  harmful  or  beneficial  effects  of 
any  types  of  illuminating  installations  will  carry  considerable 
weight.  They  have  discussed  time  after  time  how  they  were  to 
make  the  tests,  and  this  paper  has  grown  out  of  Dr.  Ferree's 
experiments,  in  the  hope  of  placing  in  the  hands  of  that  Com- 
mittee means  of  making  the  tests.  We  are  extremely  fortunate 
in  having  the  first  public  report  of  this  method.  The  method  is 
under  advisement  by  the  Committee  and  I  was  present  at  one  of 
their  meetings  when  Dr.  Ferree  brought  a  set  of  this  apparatus 


60  TRANSACTIONS  I.  E.  S. — PART  II 

to  Philadelphia  and  they  made  a  test  of  it.  Two  or  three  of  the 
doctors  present  were  very  much  impressed  with  it.  I  think,  even 
if  it  does  not  succeed  as  it  now  stands,  perhaps  here  is  the  germ 
of  the  best  possible  method  of  test. 

Dr.  C.  E.  FerrEE  (communicated  in  reply)  :  I  can  express 
only  great  appreciation  of  the  interest  that  the  men  who  have  pre- 
ceded me  have  taken  in  our  work.  The  problem  is  extremely  in- 
teresting to  me  and  I  hope  we  have  here  a  vulnerable  point  of 
attack.  Once  we  have  procured  a  successful  method  of  measur- 
ing the  effect  of  different  lighting  systems  on  the  eye,  a  broad  field 
of  application  opens  out  before  us.  We  not  only  can  find  out 
what  are  the  favorable  and  what  are  the  unfavorable  features  in  a 
lighting  system,  but  we  can  no  doubt,  as  may  be  inferred  from  Dr. 
Ives'  discussion,  test  out  and  perfect  a  lighting  system,  so  far  as 
its  effect  on  the  eye  is  concerned,  before  we  put  it  on  the  market. 
This  latter  point  is  a  good  one,  I  think,  and  I  thank  Dr.  Ives  for 
the  suggestion.*  I  feel  that  Dr.  Ives'  perspective  and  practical 
grasp  of  the  situation  is  a  distinct  contribution  to  the  paper. 

We  are  very  much  handicapped  at  present  for  funds  by  means 
of  which  to  carry  on  this  work.  In  the  first  place  apparatus  and 
models  of  lighting  systems  are  required  for  the  work  on  the 
laboratory  side.  Trained  assistants  are  also  needed  to  help  out 
with  the  details  of  the  work.  Further,  to  verify  and  enlarge 
the  work  done  in  miniature  in  the  laboratory,  we  should  test  the 
eyes  of  employees  working  under  established  lighting  systems 
and  in  the  surroundings  in  which  these  systems  have  to  operate. 
All  of  this  takes  time  and  money,  also  entrance  into  commercial 
concerns.  In  all  of  these  regards  we  need  the  help  and  influence 
of  the  Illuminating  Engineering  Society. 

This  work,  I  suppose,  could  be  done  spontaneously  and  sporadi- 
cally here  and  there  as  the  insight  and  inclination  of  various  men 
may  direct.  But  in  the  beginning,  at  least,  I  do  not  think 
it  should  be  scattered.  Until  launched  and  safely  moving,  it 
should  be  done  under  common  supervison. 

*  The  general  idea  that  over  and  above  its  application  to  abstract  investigation  the 
test  may  have  an  application  in  the  daily  work  of  the  lighting  engineer  has  come  to  the 
writer  by  suggestion  from  the  engineers  themselves.  Mr.  Cravath,  for  example,  has  re- 
cently pointed  out  that  the  test  should  be  of  advantage  in  making  the  actual  installation 
of  a  lighting  system.  The  writer  would  suggest  in  addition  that  it  may  further  be  of  ser- 
vice in  determining  the  effect  of  different  kinds  of  type  and  paper  on  the  efficiency  of  the 
eye:  also  the  effect  of  different  kinds  of  desk  lighting,  etc.  In  short,  it  is  obvious  that  the 
usefulness  of  such  a  test  is  limited  along  these  lines  only  by  its  sensitivity. 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 

Published  on  the  2Sth  of  each  month,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices.-  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIM 


FEBRUARY,  1913 


No.  2 


Index  for  Volume  VII. 

The  index  for  Volume  VII  (1912)  of 
the  Transactions  is  mailed  with  this 
number. 


Council  Notes. 

The  council  held  a  regular  meeting  in 
the  general  offices  of  the  society,  29 
West  39th  Street,  New  York,  February 
14.  1913.  Those  in  attendance  were: 
Preston  S.  Millar,  president;  George  S. 
Barrows,  Louis  Bell,  C.  O.  Bond,  J.  R. 
Cravath,  Joseph  D.  Israel,  general  sec- 
retary; V.  R.  Lansingh,  Norman  Mac- 
beth, L,.  B.  Marks,  treasurer;  W.  Cullen 
Morris,  C.  J.  Russell  and  W.  J.  Serrill. 

A  monthly  report  on  the  membership 
and  the  receipts  and  expenses  was  re- 
ceived from  the  general  secretary.  The 
number  of  members,  counting  the 
applications  and  resignations  presented 
at  the  meeting,  was  said  to  total  1,325. 

Eleven  applicants  were  elected  mem- 
bers. Their  names  appear  on  another 
page  of  this  number. 

Reports  on  section  activities  were  re- 
ceived from  Vice-presidents  J.  R. 
Cravath  (Chicago),  Howard  S.  Evans 
(Pittsburgh).  J-.  W.  Cowles  (Boston), 
\V.  J.  Serrill    (Philadelphia). 

A  tentative  report  on  proposed  work 
was  received  from  the  chairman  of  the 
section  development  committee.  It  was 
stated  that  the  first  meeting  of  the  com- 


mittee was  to  be  held  February  15  and 
that  a  more  definite  report  would  be 
presented   later. 

An  oral  report  was  received  from  the 
papers  committee.  The  arrangements 
for  the  papers  program  of  the  next  con- 
vention were  discussed  briefly.  It  was 
said  that  it  might  be  well  to  arrange 
the  next  convention  program  somewhat 
as  follows :  the  first  day  to  be  devoted 
to  society  affairs;  the  second  day  to 
the  presentation  of  technical  papers; 
the  third  day  to  the  reading  of  papers 
of  a  commercial  character;  and  the  last 
day  to  a  series  of  lectures  and  talks  on 
the  architectural  and  decorative  aspects 
of  illuminating  engineering.  It  was 
understood  that  a  definite  report  would 
be  received  from  the  papers  committee 
at  the  March  meeting  of  the  council. 

In  a  report  received  from  the  com- 
mittee on  editing  and  publication  it  was 
stated  that  beginning  with  the  1913 
issue  of  the  Transactions  an  inexpen- 
sive mat  surface  paper  for  the  ordinary 
text  matter  and  line  illustrations  and 
a  coated  paper  for  the  photographic 
reproductions  would  be  used ;  each  issue 
of  the  Transactions  will  be  printed  in 
two  parts;  Part  I  will  be  devoted  to 
council  and  section  notes  and  news 
items;  Part  II  will  include  papers,  dis- 
cussions and  important  committee  re- 
ports. The  committee  also  asked  for 
an  appropriation  of  $20.00  to  print  two 
or    three    hundred    copies    of    a    guide 


TRANSACTIONS   I.    K.    S. — PART   I 


setting  forth  the  character  and  style  of 
papers  for  presentation  at  meetings  of 
the  society  and  for  publication  in  the 
Transactions.  The  appropriation  was 
authorized. 

Progress  reports  were  received  from 
the  sustaining  membership  committee, 
the  committee  on  new  membership  and 
the  committee  on  reciprocal  relations 
with  other  societies. 

A  tentative  report  from  the  commit- 
tee on  glare  from  reflecting  surfaces 
outlined  the  work  of  the  committee  for 
the  present  year.  In  its  work  the  com- 
mittee will  give  special  attention  to 
school  officials  and  school  book  com- 
panies. The  co-operation  of  other 
societies  will  be  solicited.  Information 
regarding  paper  making  and  the  pub- 
lishing art  will  be  continually  sought. 
Encouragement  will  be  given  to  manu- 
facturers to  produce  a  dull  finished 
paper  which  will  reproduce  half-tones 
satisfactorily.  The  committee  also  be- 
lieves that  it  is  possible  to  collect  suffi- 
cient data  regarding  the  use  of  black- 
boards, polished  desk  tops,  glazed  paper, 
etc.,  to  form  a  paper  of  sufficient  value 
to  find  a  place  on  the  program  of  the 
next  convention  of  the  society.  Briefly 
speaking,  the  committee  will  make  every 
effort  toward  the  elimination  of  polished 
surfaces,  but  for  the  present  it  will 
confine  its  efforts  to  the  public  schools. 

The  finance  committee  reported  that 
it  had  approved  for  payment  vouchers 
Nos.  1170  and  1173,  inclusive,  and  1175 
to  1209,  inclusive,  aggregating  $759.02. 

The  factory  lighting  legislation  com- 
mittee reported  that  the  recommenda- 
tions which  it  had  submitted  to  the 
New  York  Factory  Investigating  Com- 
mission pertaining  to  bill  No.  18  having 
to  do  with  the  lighting  of  factories  and 
work  rooms  had  been  incorporated  in 
the  revised  bill  of  the  commission. 


An  oral  report  was  received  from 
the  committee  on  illumination  primer. 
The  committee  asked  that  the  council 
authorize  the  publication  of  a  large  edi- 
tion of  the  primer  to  fill  any  orders  that 
may  be  received,  and  that  permission 
be  given  to  make  a  few  changes  in  the 
primer.  The  council  authorized  the 
committee  to  make  such  changes  as  it 
may  deem  necessary,  and  the  printing 
of  such  editions  of  the  primer  as  may 
be  required  from  time  to  time.  It  was 
understood  that  these  editions  would 
be  gotten  out  by  the  general  office. 

A  report  was  received  from  Mr.  C.  A. 
Littlefield,  chairman  of  the  committee 
of  arrangements  for  the  1913  annual 
meeting.  It  was  resolved  to  extend  to 
Mr.  Littlefield  and  the  members  of  his 
committee,  on  behalf  of  the  society,  a 
very  hearty  vote  of  thanks  for  the 
excellent  and  successful  meeting  and 
dinner  which  they  had  arranged. 

Upon  receipt  of  a  report  from  Mr. 
V.  R.  Lansingh,  chairman  of  a  pre- 
liminary committee,  it  was  resolved  that 
the  president  appoint  a  committee  to 
foster  the  establishment,  maintenance 
and  development  of  courses  in  illumi- 
nating engineering  in  colleges  and  uni- 
versities. The  president  appointed  the 
following  committee,  which  is  to  be 
known  as  the  committee  on  collegiate 
education :  V.  R.  Lansingh,  chairman ; 
Prof.  Henry  B.  Dates,  secretary;  Df. 
H.  E.  Ives. 

The  president  was  requested  to  ap- 
point a  committee  to  consider  and  re- 
port to  the  council  upon  the  advisability 
(1)  of  holding  the  1915  annual  con- 
vention in  San  Francisco,  and  (2)  of 
endeavoring  to  arrange  for  that  time 
and  place  a  joint  meeting  of  the  several 
illuminating  engineering  societies  or  a 
meeting  of  the  proposed  International 
Commission  on  Illumination,  or  both. 


TRANSACTIONS    I.    E.    S. — PART    I 


The  president  was  also  requested  to 
appoint  a  committee  on  popular  lectures, 
said  committee  to  be  asked  to  promul- 
gate a  plan  for  preparation  of  popular 
lectures  on  various  classes  of  lighting 
installations  (as  factory,  store,  resi- 
dence, etc.)  and,  after  plan  has  been 
approved  by  the  council,  to  undertake 
the  preparation  of  such  lectures,  either 
directly   or   through   sub-committees. 

It  was  resolved  that  the  council  of 
the  Illuminating  Engineering  Society 
takes  pleasure  in  accepting  the  invita- 
tion of  the  American  Gas  Institute  to 
join  in  a  session  on  illumination  at 
Richmond  Ya.,  on  October  16,  1913. 


New  Members. 

The  following  applicants  were  elected 

members  of  the  society  at  a  meeting  of 

the  council  held  February   14,   1913 : 

Burrows,  S.  B. 

Public  Service  Electric  Co.,  Newark, 
N.  J. 

Clinch,  Edward  S.,  Jr. 

Lord  Electric  Co.,  105  West  40th 
Street,  New  York,  N.  Y. 

Frey,  Arthur  C. 

Cadet  Engineer,  United  Gas  Im- 
provement Co..  Broad  and  Arch 
Streets,    Philadelphia,    Pa. 

Gudge,  B.  J. 

Electrical  Engineer.  Westinghouse 
Elec.  &  Mfg.  Co..  East  Pittsburgh, 
Pa. 

Higgixs,  Warren  Snedek. 

Instructor  in  Electrical  Engineering, 
Lafayette  College,  Easton,   Pa. 

Horner,   Harry  Archer. 

Electrical  Engineer.  New  York 
Shipbuilding  Co.,  Camden,  N.  J. 

Kelsey,  Fenton   P. 

Vice-President  and  Editor.  Gas 
Record,  Chicago,  111. 


Malia,  James  P. 

Chief  Electrician,  Armour  &  Co., 
Union  Stock  Yards.  Chicago,  111. 

McCulloch,  Fred.  H. 

Treasurer,  Electric  Supply  &  Fix- 
ture Co.,  123  East  Washington 
Street,   Fort   Wayne.   Ind. 

Skinner.  Ross  H. 

Assistant  to  General  Contracting 
Agent,  Consolidated  Gas  Co.,  435 
Sixth  Avenue,  Pittsburgh,  Pa. 

Stafford,  Raymoxd  W. 

Assistant  Foreman,  The  New  York 
Edison  Company,  117  West  39th 
Street,  New  York  City. 


Section  Activities. 

CHICAGO  SECTIOX 

A  joint  meeting  of  engineering,  archi- 
tectural and  ophthalmological  societies 
was  held  at  the  Republican  House  in 
Milwaukee,  February  22,  1913.  The 
following  societies  participated : 

Chicago  Section  of  the  Illuminating 
Engineering  Society. 

Engineering   Society  of   Wisconsin. 

Milwaukee  Company  Section  of  the 
National  Electric  Light  Association. 

Milwaukee   Electrical   League. 

Milwaukee   Engineering   Society. 

Milwaukee   Oto-ophthalmic   Club. 

Milwaukee  Section  of  the  American 
Chemical  Society. 

Milwaukee  Section  of  the  American 
Institute  of  Electrical  Engineers. 

Madison  Section  of  the  American 
Institute  of  Electrical  Engineers. 

Wisconsin  Chapter  of  the  American 
Institute  of   Architects. 

The  program  presented  was  as 
follows; 

"Light  and  Art,"  by  Mr.  M.  Luckiesh, 
engineer,  National  Electric  Lamp  Asso- 
ciation, Cleveland,  O. 

Discussion  of  "Ocular  Comfort  and 
its    Relation   to   Glare   from    Reflecting 


TRANSACTIONS    I.    E.    S. — PART    I 


Surfaces,"  by  Mr.  F.  A.  Vaughn,  con- 
sulting engineer,  and  Dr.  Nelson  M. 
Black,  ophthalmologist,  Milwaukee, 
Wis. 

"A  Photometer  Screen  for  Use  in 
Tests  of  Street  Illumination,"  by  Prof. 
Arthur  H.  Ford,  State  University  of 
Iowa,  Iowa  City. 

"The  Influence  of  Colored  Surround- 
ings on  the  Color  of  the  Useful  Light," 
by  Mr.  M.  Luckiesh,  engineer,  National 
Electric  Lamp  Association,  Cleveland, 
O. 

These  papers  and  discussions  will 
appear  in  later  issues  of  the  Trans- 
actions. The  meeting  was  well  at- 
tended, some  30  or  40  Chicago  men, 
besides  members  from  Madison,  Mil- 
waukee and  other  Wisconsin  and  Michi- 
gan points  being  present. 

NEW  ENGLAND   SECTION 

The  New  England  section  held  a  joint 
meeting  with  the  Boston  section  of  the 
American  Institute  of  Electrical  Engi- 
neers in  the  auditorium  of  the  Boston 
Edison  Building,  39  Boylston  Street, 
Boston,  February  17,  1913.  The  follow- 
ing papers  were  read : 

"Street  Lighting  with  Ornamental 
Luminous  Arc  Lamps"  by  C.  A.  B.  Hal- 
vorson,  Jr.,  of  the  General  Electric  Com- 
pany, West  Lynn,  Mass. 

"The  Theory  of  Mercury- Vapor  Ap- 
paratus" by  P.  H.  Thomas,  consulting 
engineer,  New  York. 

"Flame  Carbon  Arc  Lamps"  by  W.  A. 
Darrah  of  the  Westinghouse  Electric 
&  Manufacturing  Company,  East  Pitts- 
burgh, Pa. 

The  first  two  of  these  papers  appear 
in  this  issue  of  the  Transactions.  The 
last  one  will   be  published   later. 

NEW    YORK    SECTION 

A  joint  meeting  of  the  New  York 
section  of  the  I.  E.  S.  and  the  Munici- 


pal Art  Society  was  held  in  the  National 
Arts  Club,  119  East  19th  Street,  New 
York,  February  12,  1913.  The  topic  of 
the  evening  was  "Municipal  Lighting." 
Addresses  were  made  by  C.  F.  Lacombe, 
chief  engineer  of  the  Department  of 
Water  Supply,  Gas  and  Electricity  of 
New  York  City;  Charles  Roland  Lamb, 
ecclesiastical  architect;  William  Wentz, 
vice-president,  O.  J.  Gude  Company,  and 
Arthur  Williams,  general  inspector,  The 
New  York  Edison  Company.  A  dinner 
preceded  the  meeting  and  was  attended 
by  guests  and  members  of  both  the 
club  and  the  society. 

The  program  of  meetings  for  the  re- 
mainder of  the  season  is  as  follows : 

March  13 — Joint  meeting  with  the 
American  Society  of  Mechanical  Engi- 
neers in  the  United  Engineering  So- 
cieties Building,  29  West  39th  Street, 
New  York.  Mr.  Ward  Harrison  of  the 
National  Electric  Lamp  Association  will 
present  a  paper  on  "Industrial  Lighting." 

April  9 — This  meeting  will  probably 
be  held  in  the  United  Engineering 
Societies  Building.  Mr.  M.  Luckiesh  of 
the  National  Electric  Lamp  Association 
will  present  a  paper  on  "Light  and  Art." 
A  paper  on  "Phosphorescence  and 
Fluorescence"  is  also  scheduled.  This 
meeting  should  be  an  unusually  inter- 
esting one. 

May  8 — A  talk  on  theater  lighting  by 
Mr.  Bassett  Jones,  Jr.,  at  the  Clymer 
Street  Theater,  Brooklyn.  During  the 
past  year  Mr.  Jones  has  conducted  a 
great  deal  of  experimental  work  in 
theater  illumination  particularly  in  the 
production  of  stage  effects.  The  mem- 
bers of  the  New  York  chapter  of  the 
American  Institute  of  Architects  will  be 
invited  to  attend  this  meeting.  Admis- 
sion will  be  by  card. 

June  8 — It  is  planned  to  have  a  joint 
meeting  and  outing  of  all  the  engineer- 
ing societies  in  New  York. 


TRANSACTIONS    I.    E.    S.  —  PART    I 


PHILADELPHIA    SECTION 

The  February  meeting  of  the  Phila- 
delphia section  of  the  Illuminating  En- 
gineering Society  was  held  on  Feb- 
ruary 21,  1913,  at  the  Engineers'  Club, 
1317  Spruce  Street,  Philadelphia.  At 
the  dinner  preceding  the  meeting,  which 
was  held  in  the  dining-room  of  the 
Engineers'  Club,  46  members  and  guests 
were  present.  One  hundred  and  sev- 
enty-five members  including  a  number 
of  architects  and  ophthalmologists  were 
present. 

Mr.  P.  S.  Millar,  president  of  the 
society,  presented  a  paper  on  "Some 
Phases  of  .the  Illumination  of  Inte- 
riors." The  subject  was  discussed  by 
Messrs.  Perot,  Bond,  Regar,  Hare, 
Dickey,  Swanfeld.  Israel,  Prof.  Hoadley 
and  Dr.  Crampton. 


PITTSBURGH     SECTION 

A  meeting  of  the  Pittsburgh  section 
was  held  in  the  hall  of  the  Engineers' 
Society  of  Western  Pennsylvania,  Oliver 
Building  Pittsburgh,  February  21,  1913. 
A  paper  on  gas  illumination  by  Mr. 
S.  B.  Stewart  of  the  Consolidated  Gas 
Company  was  presented. 

The  program  of  meetings  for  the  rest 
of  the  season  is  as  follows  : 

March — "Moving  Picture  Lanterns 
from  the  Central  Station  Point  of 
View"  by  J.  F.  Martin. 

April — "Railroad  Car  Lighting"  by 
J.  L.  Minick. 

May — "Physiological  Aspects  of  Il- 
lumination" by  W.  E.  Reed. 

June — Announcement  will  be  made 
later. 


to 


TRANSACTIONS 


OF  THE 


Illuminating 
Engineering  Society 

FEBRUARY,  1913 
PART  II 

Papers,  Discussions  and  Reports 


[  FEBRUARY,  1913  ] 

CONTENTS  -  PART  II 


The  Influence  of  Colored  Surroundings  on  the  Color  of  the 
Useful  Light.     By  M.  Luckiesh 61 

Theory  of  Mercury-vapor  Apparatus.     By  Percy  H.  Thomas  75 

Street  Lighting  with  Ornamental  Luminous  Arc  Lamps.  By 
C.  A.  B.  Halvorson,  Jr SS 


THE  INFLUENCE  OF  COLORED  SURROUNDINGS  ON 
THE  COLOR  OF  THE  USEFUL  LIGHT.* 


BY    M.    LUCKIESH. 


Synopsis :  In  the  course  of  the  investigation  reported  in  this  paper, 
theoretical  computations  and  also  actual  colorimeter  measurements  were  made 
to  determine  the  magnitude  of  the  influence  of  colored  surroundings  on  the 
color  of  the  useful  light.  Much  attention  has  heretofore  been  given  to  the  colo  r 
value  of  illuminants,  when  after  all  the  color  value  of  the  useful  light  is  per- 
haps of  greater  interest  to  the  user.  It  was  suspected  that  especially  in  in- 
direct lighting  the  colored  walls  and  ceiling,  even  though  nearly  white, 
would  effect  appreciable  color  changes  in  the  incident  light.  Calculations 
given  in  this  paper  show  that  only  a  moderately  yellow  paper  is  required  to 
change  the  color  of  tungsten  light  to  that  of  a  carbon  filament  lamp.  Both 
direct  and  indirect  lighting  were  considered  in  the  computations  ;  the  results 
obtained  are  plotted  in  several  ways. 

A  miniature  room  was  fitted  with  a  tungsten  unit  in  the  middle  of  the 
ceiling.  The  walls  and  ceiling  were  covered  with  combinations  of  green- 
yellow,  and  white  paper.  Colorimeter  readings  were  made  with  direct  and 
indirect  lighting.  These  are  shown  in  a  table  and  in  a  color-triangle.  Re- 
flection coefficients  are  briefly  discussed  and  measurements  presented  for 
several  papers  illuminated  by  various  illuminants.  These  measurements 
were  made  with  a  flicker  photometer. 

The  color  value  of  illuminants  has  been  a  subject  of  consider- 
able interest  during  the  last  few  years.  The  light  from  the 
modern  metallic  filament  lamps  has  been  lauded  by  many  because 
of  its  nearer  approach  to  daylight  than  many  of  the  other  common 
illuminants,  while  a  few  have  expressed  favor  for  the  "softer" 
yellow  light  of  the  carbon  filament  lamp.  Relatively  the  color 
values  of  common  illuminants  differ  considerably.  When  these 
illuminants  are  used  amid  colored  surroundings  one  is  led  to 
suspect  that  considerable  change  in  the  color  value  of  the  useful 
light  might  take  place  especially  with  indirect  and  semi-indirect 
lighting.  Of  course  great  color  changes  can  be  effected  by  using 
colored  reflecting  surfaces,  the  amount  of  alteration  in  color 
depending  upon  conditions.  For  instance,  there  is  no  theoretical 
reason  why  the  light  from  ordinary  illuminants  cannot  be  altered 
by  reflection  from  colored  surfaces  to  accurately  match  "average 

1  A  paper  read  at  a  meeting  of  the  Chicago  section  of  the  Illuminating:  Engineering 
Society,  in  Milwaukee,  Wis.,  February  22,  1913. 


62  TRANSACTIONS  I.  E.  S. — PART  II 

daylight."  In  nature  considerable  change  takes  place  in  the  color 
of  sunlight  on  being  reflected  from  the  earth  and  especially  from 
green  foliage.  This  has  been  shown  by  noting  in  this  reflected 
light  the  absorption  bands  of  chlorophyl  which  is  the  green  color- 
ing matter  of  foliage.  Measurements  at  sea  or  over  a  sufficiently 
large  snOw  field  in  winter  do  not  show  these  peculiarities.  Mr. 
G.  S.  Merrill1  has  measured  the  color  values  of  daylight  on  the 
working  plane  in  a  room  after  part  of  it  had  been  reflected  from 
the  colored  surroundings.  The  interior  measurements  were  made 
on  clear  and  cloudy  days.  They  showed  considerable  alteration 
in  the  color  of  outdoor  average  daylight. 

In  direct  artificial  lighting  much  less  light  reaches  the  working 
plane  via  the  walls  and  ceiling  than  in  semi-indirect  or  indirect 
lighting.  Obviously  in  the  latter  system  all  the  light  which  is 
diffusely  reflected  by  colored  walls  or  ceilings  is  altered  in  color. 
When  the  reflection  coefficient  is  small  the  influence  of  the  colored 
paper  is  small,  especially  in  direct  lighting  owing  to  the  small 
amount  of  the  colored  light  which  is  added  to  the  direct  light. 
In  such  cases  only  about  two  reflections  need  be  considered,  as 
there  is  little  light  remaining  unabsorbed  after  it  has  suffered  two 
reflections.  When  the  reflection  coefficient  of  the  surroundings 
is  large  many  more  reflections  must  be  taken  into  account  as  will 
be  shown  later. 

Assume  an  illuminant  radiating  equal  amounts  of  monochro- 
matic red,  green  and  blue  light  which  might  be  represented 
relatively  as 

RGB 

IOO  IOO  IOO 

Further  assume  that  this  light  source  is  placed  in  the  center  of 
an  Ulbricht  sphere,  the  walls  of  which  are  covered  with  perfectly 
diffusing  green  paper.  The  reflection  coefficients  of  this  paper 
for  the  particular  illuminant  are  assumed  to  be 

R  g  b  , 

25.2  47-2  27.6 

The  light  received  by  the  green  paper  is  reflected  an  infinite 
number  of  times.    If  the  walls  of  the  sphere  are  temporarily  sup- 

1  Proc.  American  Institute  of  Electrical  Engineers,  p.  1726,  1910. 


LUCKIESH  :      INFLUENCE  OF  COLORED   SURROUNDINGS  63 

posed  to  be  white  and  if  N  is  the  reflection  coefficient  of  the 
paper  then  the  total  light  falling  on  the  walls  will  be 

Q  =  Q'  +  NQ'  +  N'Q'  +  N8Q'  +  =--  y^-        (1) 

where  Q  =  total  light  falling  on  the  walls, 

Q'  =  direct  light  from  the  light  source  falling  on  the 
walls. 

Color  values  of  a  paper  are  determined  by  measuring  the  color 
value  of  the  light  after  it  has  suffered  one  reflection.  A  reflec- 
tion coefficient  of  33//i  per  cent,  was  assumed  for  the  green  paper 
for  this  particular  illuminant.  The  coefficient  of  reflection  of 
the  paper  can  vary  between  wide  limits  without  any  change  in 
the  color  values. 

Based  on  the  foregoing  assumptions,  the  reflection  coefficient 
of  this  paper  for  the  monochromatic  red  light  is  25.2  per  cent, 
of  the  original  100  units;  47.2  per  cent,  of  the  total  100  units  of 
green  light;  27.6  per  cent,  for  the  total  100  units  of  blue  light. 
For  this  case  the  total  red,  green  and  blue  components  in  the 
light  incident  on  the  wall  paper  after  an  infinite  number  of  reflec- 
tions will  be 

QR  =  Q'R+NRQ'R+NK?Q'R+N|iO'R+  . . .  =       Q\T         (2) 


Qo  =  Q'c  +  NcQ'c  +  Ng-Q'g  +  Ng3Q'g  +  •  •  •  =  t    ^V       (3) 


I 

—  NR 

Q'c 

I 

-NG 

Q'b 

Qb  =  Q'b  +  NBQ'B  +  NJQ'b  +  N|Q'B  +  .  .  .  =  t  ^^  (4) 
and 

Q  =  Qr  +  Qo  +  Qb  =  total  light  on  walls  (5) 

Q'  =  Q'r  +  Q'c  +  Q'b  =  total  direct  light  on  walls  (6) 

NR,  NG,  NB  are  respectively  the  reflection  coefficients  for  the 
monochromatic  red,  green  and  blue  light's. 

NRQ'R,  NGQ'G,  NbO'b  are  the  color  values  of  the  wall  paper  as 
determined  by  the  colorimeter  under  the  light  Q\ 

It  is  interesting  to  make  some  calculations  in  this  particular 
case. 


64 


TRANSACTIONS  I.  E.  S. — PART  II 


TABLE  I.— Computations  According  to  Equations  (2),  (3)  and  (4), 
Showing  the  Changes  Produced  in  the  Light  from  a  Special 
"White  Source"  by  Successive  Reflections  from  a  Certain 
Green  Paper  a. 


The  term  in 

Values 

Values  reduced  for  r. 
color  trianj 
R               G 

lotting  in 

(2),  (3)  and  (4)          R 

G 

B 

B 

0/ 

100 

IOO 

IOO 

33-3 

33-3 

33-3  a 

NQ/.-- 

25.20 

47.20 

27.60 

25.2 

47-2 

27.6  b 

N2Q/  •  ■  • 

6-35 

22.30 

7.62 

17-5 

61.5 

21.0  c 

N3Q/.-. 

1.60 

IO.45 

2.IO 

n-3 

73-9 

14.8  d 

NV- 

0.40 

4-93 

O.58 

6.8 

83.4 

4.8  e 

N50/  •  ■  ■ 

0. 10 

2-33 

O.I6 

3-8 

90.0 

6.2  f 

N6Q/  • • • 

0,03 

1. 10 

O.04 

2.6 

94.0 

3-4  g 

N7Q/--- 

0.52 

O.OI 

N8Q/  •  ■  • 

0.25 

N9Q/  •  • • 

0.12 

The  reduced  values  a,  b,  c,  etc.,  are  plotted  in  fig.  1.    Theoret- 
ically these  values  should  be  plotted  in  the  color  pyramid  as  they 


Fig.  1.— Theoretical  computations  of  the  changes  taking  place  in  the  color  of  light 

under  various  conditions  and  after  undergoing  various  reflections 

are  plotted  in  trilinear  coordinates. 

are  of  unequal  luminosity  but  for  clearness  they  are  plotted  in 
one  plane — the  color  triangle — which  method  shows  their  rela- 


luckiesh:     influence  of  colored  surroundings 


tive  positions.  It  will  be  noted  how  the  color  of  the  light 
approaches  saturated  green  as  the  number  of  reflections  increases. 
In  fig.  2  are  plotted  the  actual  luminosity  values  of  the  red,  green 
and  blue  components  in  the  light  after  it  has  undergone  various 
reflections.  The  full  lines  represent  the  magnitude  of  the  com- 
ponents when  the  reflection  coefficient  of  the  paper  is  ZVA  Per 
cent,  for  the  illuminant  used.  The  dotted  lines  show  the  rapid 
decrease  in  the  values  with  a  paper  of  10  per  cent,  reflection 
coefficient  as  compared  with  one  of  335/3  per  cent,  coefficient.  This 
data  of  course  is  not  truly  represented  by  a  curve  because  there 


100 

90 
80 

</> 

560 

t— 

z 

£50 
J40 

UJ 

ac 
30 

20 

10 


i\ 

in  1  \ 

1  vU  \ 

'  \\\  \ 

1  \\\  \ 
1  \\\  \ 

| 

1  v  > 

1 

v  \ '* 

\      \ 

\ 

^R^N 

-  -~r 

s5s**>i~- 

___ 

D 

• 

>                  ; 

' 

1 

>                  ( 

NUMBER  0FREFtECTIONS,6REEN  PAPER 

Fig.  2.— The  percentage  of  the  original  light  remaining  after  it  has 
suffered  various  reflections. 

cannot  be  fractional  parts  of  a  reflection.  The  various  points 
corresponding  to  different  reflections  are  connected  by  the  curves 
merely  as  a  graphical  means  of  representing  the  data.  In  fig.  3 
are  plotted  the  relative  amounts  of  the  red,  green  and  blue  com- 
ponents in  the  light  after  it  has  suffered  various  reflections.  The 
sum  of  the  ordinates  at  any  particular  reflection  equals  100  per 
cent.  Here  the  rapid  approach  toward  a  pure  green  is  shown 
after  the  light  has  undergone  several  reflections. 

Turning  again  to  fig.  1  and  Table  I  some  more  interesting  cal- 
culations are  possible.  First  let  us  discard  the  direct  light  from 
the  calculations.  Adding  the  light  of  the  first  and  second  reflec- 
tions in  columns  2.  3  and  4  and  making  the  sum  of  the  three 


66 


TRANSACTIONS  I.  E.  S. — PART  II 


components  equal  to  ioo  we  have  the  quality  of  the  light  reflected 
once  from  the  paper  but  diluted  with  the  light  of  the  second 
reflection.  This  latter  amount  is  small  but  more  greenish  in  color. 
This  value  is  plotted  with  a  cross  and  numbered  2,  which  indi- 
cates the  sum  of  two  reflections.  The  sum  of  three,  four  and 
five  reflections  are  also  plotted  as  crosses  and  labeled  consistently. 
The  final  color  of  the  light  after  undergoing  an  infinite  number 
of  reflections  (and  absorptions)  is  found  by  summating  the 
series  in  equations  2,  3  and  4  and  subtracting  in  each  case  the 
direct  component.     This   value   for   indirect  lighting  might  be 


100 
so 

£0 
70 
60 
50 

30 
£0 
10 

°< 

e""--4/ 

B"\ 

R--' 

) 

r 

NUM 

3ER0FREE 

5                  i 
.ECTIONS.G 

REEN  PAPE 

>                  6 
R 

Fig.  3. — The  percentage  of  the  three  components  in  the  light  after  undergoing 
various  reflections  frorn  green  paper. 

found  by  measuring  the  color  of  a  white  surface  screened  from 
the  light  source.     It  is  plotted  as  °o  . 

Next  the  effect  of  adding  a  constant  amount  of  direct  light  Q' 
to  the  reflected  light  is  shown  by  the  circles.  Circle  number  i' 
shows  the  color  of  the  direct  light  as  diluted  by  the  first  reflec- 
tion. Circles  2' ',  3',  4',  6"  indicate  the  change  in  the  color  of  the 
total  light  as  more  reflections  are  considered.  The  color  of  the 
total  light  after  it  has  undergone  an  infinite  number  of  reflections 
is  found  by  summating  each  series  in  equations  2,  3  and  4,  and 
making  the  sum  of  the  three  values  equal  to  100.  It  is  plotted 
as  00 '. 


LUCKIESH  :      INFLUENCE  OF   COLORED   SURROUNDINGS  67 

After  purely  theoretical  computations  made  only  for  the  pur- 
pose of  showing  the  order  of  magnitude  of  the  possible  color 
changes,  it  becomes  of  interest  to  make  some  observations 
with  an  Ives  colorimeter  and  in  a  room  papered  with  various 
colored  wall  papers.  With  actual  colored  papers  as  used  on 
walls  there  is  more  or  less  specular  reflection.  The  light  which 
is  specularly  reflected  does  not  necessarily  undergo  the  same 
change  in  its  color  as  in  the  case  of  diffuse  reflection.  This  fact 
must  be  considered  when  using  the  colorimeter  for  determining 
the  color  of  papers  and  other  things.  If  the  paper  is  placed  at 
such  an  angle  as  to  regularly  reflect  the  light  from  its  illuminant 
into  the  colorimeter  obviously  the  true  color  of  the  paper  will 
not  be  obtained.  Mere  observation  indicates  that  more  likely  the 
colorimeter  readings  will  more  nearly  represent  the  color  of  the 
illuminant  than  that  of  the  paper. 

It  is  impractical  to  use  an  actual  room  in  these  experiments 
owing  to  the  great  amount  of  light  required  for  the  colorimeter 
readings  and  also  the  large  surfaces  which  must  be  re-papered. 
Besides  the  results  would  only  hold  for  that  particular  room,  wall 
paper,  etc.  This  of  course  is  true  with  any  arrangement  of 
apparatus.  The  object  of  these  measurements  was  merely  to 
show  the  possible  magnitude  of  the  color  changes  in  the  illumi- 
nant after  it  had  suffered  various  reflections  and  had  reached  the 
working  plane.  A  cubical  box,  four  feet  on  a  side,  was  arranged 
with  a  single  fixture  in  the  center  of  the  ceiling.  The  light  source 
was  a  500-watt  tungsten  lamp.  This  source  was  used  directly 
and  indirectly.  With  the  direct  system  no  reflector  was  used,  of 
course  permitting  considerably  more  than  the  usual  percentage  of 
light  flux  to  reach  the  walls  and  ceiling.  Green,  yellow  and  white 
papers  were  used.  The  green  and  yellow  papers  were  selected 
from  regular  wall  paper  stock  and  were  quite  unsaturated  colors. 
The  white  paper  used  was  blotting  paper.  All  colorimeter  read- 
ings were  made  with  the  photometric  field  of  constant  brightness. 
The  colorimeter  readings  under  various  conditions  are  shown 
in  Table  II.  It  will  be  noted  that  the  readings  for  the  tungsten 
lamp  are  shown  as  being  equal  for  the  red,  green  and  blue.  This 
course  was  considered  legitimate  for  several  reasons.  First  it  was 
found  impossible  to  make  the  readings  comparable  with   data 


68  TRANSACTIONS  I.  £.  S. — PART  II 

heretofore  published  by  others  without  entailing  a  vast  amount  of 
work  which  was  considered  unwarranted  owing  to  the  fact  that 
the  cases  are  not  general.  Further  the  chief  object  was  to  show 
only  the  order  of  magnitude  of  the  color-changes  compared  with 
the  magnitude  of  the  color  difference  between  the  carbon  and 
tungsten  incandescent  lamps.  The  results  are  plotted  in  trilinear 
co-ordinates  in  fig.  4.  It  is  interesting  to  note  that  4,  5,  6,  7  show 
a  gradual  change  from  yellow  toward  "white"  as  would  be  ex- 
pected in  going  from  indirectly  lighted  yellow  walls  and  ceiling  to 
a  white  ceiling  and  direct  lighting.  These  results  are  consistent 
.  with  the  theoretical  computations  in  the  first  part  of  the  paper. 
The  color  values  of  the  carbon  lamps  (shown  as  2  and  3  in 
Table  II)  are  relative  to  the  tungsten  lamp  and  are  given  in 
order  to  illustrate  the  magnitude  of  the  change  in  the  color  of 

TABLE  II. — Colorimeter  Measurements  in  a  Miniature 
Room  under  Various  Conditions. 

RGB 

1  Tungsten  lamp 33.3  33.3  33.3 

2  Carbon  lamp,  3.1  watts  per  candle 38.7  34.7  26.6 

3  Carbon  lamp,  4  watts  per  candle 43.0  33.7  23.3 

4  Yellow  walls  and  ceiling,  indirect 53.1  37.0  9.9 

5  Yellow  walls  and  ceiling,  direct 47.6  35.7  16.7 

6  Yellow  walls  and  white  ceiling,  indirect.  •  •  43.2  35.5  21.3 

7  Yellow  walls  and  white  ceiling,  direct 42.1  35.3  27.6 

8  Yellow  paper  .  •  •  • 43.6  3S.4  18.0 

9  Green  paper 34.5  39.8  25.7 

10  Green  ceiling  and  green  walls,  indirect- ••  •  35.9  43.6  17.5 

11  Green  ceiling  and  green  walls,  direct 36.3  37.8  25.9 

12  Green  walls  and  yellow  ceiling,  indirect .  ■•  48.2  42.6  9.2 

13  Green  walls  and  yellow  ceiling,  direct 39.8  39.6  20.6 

14  Green  walls  and  white  ceiling,  indirect 36.6  34.1  29.3 

15  Green  walls  and  white  ceiling,  direct 35.2  34.6  30.2 

the  tungsten  light  due  to  colored  surroundings.  The  color  values 
of  the  yellow  and  green  papers  used  in  the  box  are  given  at  8 
and  9  and  in  the  table.  These  were  determined  with  the  colo- 
rimeter and  show  the  color  of  the  tungsten  light  after  being 
once  reflected  from  the  paper. 

An  instrument  for  observing  the  change  in  the  color  of  incan- 
descent lamp  light  due  to  colored  surroundings  is  easily  made. 
A  box  16  inches  long  and  4  inches  square  contains  a  lamp  of 
the  same  type  and  voltage  as  used  in  the  room  in  question.    This 


LUCKIESH  :       INFLUENCE  OF   COLORED   SURROUNDINGS  69 

lamp  is  fastened  in  one  end  of  the  box  and  illuminates  a  piece 
of  opal  glass.  This  portion  of  the  box  should  have  a  white  coat- 
ing on  its  interior  walls.  Variation  in  the  brightness  of  the  glass 
is  obtained  by  means  of  a  diaphragm.  In  an  adjacent  compart- 
ment is  a  mirror  inclined  at  45  deg.  to  the  vertical.  Above  this 
is  a  ground  opal  glass  which  receives  the  light  just  as  it  is 
received  on  the  working  plane.  By  means  of  the  mirror  this  glass 
is  seen  adjacent  to  the  glass  illuminated  by  the  electric  lamp. 
When  a  brightness  match  is  made  the  change  in  the  color  of  the 
illuminant  is  observable.  Such  a  box  was  made  and  observable 
changes  in  the  color  of  the  illuminant  due  to  colored  surroundings 


Fig.  4.— Results  obtained  in  a  small  room  (plotted  from  Table  II.). 

were  noted  even  in  extreme  cases  of  direct  lighting.  A  modifica- 
tion can  be  made  by  using  in  place  of  the  electric  lamp  a  ver- 
tical tube  which  admits  light  to  the  first  opal  glass  from  the 
illuminant  only  in  much  the  same  manner  as  the  Sharp-Miller 
photometer  is  used  for  candle-power  measurements.  This  latter 
arrangement  could  be  used  with  all  kinds  of  illuminants,  but 
would  be  useless  in  indirect  lighting  where  perhaps  the  greatest 
color  changes  are  found. 

Although  it  is  well  known  that  the  coefficient  of  reflection  of  a 
surface  is  not  the  same  for  all  illuminants  unless  that  surface  is 
white,  it  may  not  be  out  of  place  to  discuss  reflection  coefficients 


70  TRANSACTIONS  I.  E.  S. — PART  II 

briefly.  With  the  ever  increasing  diversity  of  types  of  light 
sources  and  refinement  of  illuminating  engineering  procedure  it 
may  soon  become  desirable  when  giving  reflection  coefficients  to 
also  include  in  the  statement  the  light  source  for  which  the  coeffi- 
cient was  determined.  To  illustrate  this  point  some  relative 
reflection  coefficients  of  various  colored  papers  of  high  saturation 
are  given  in  Table  III.  These  were  determined  only  relatively  by 
measuring  the  relative  brightnesses  in  a  fixed  direction  while  the 
specimen  was  illuminated  successively  by  various  illuminants  at 
hand.  The  brightness  measurements  were  made  by  means  of  a 
flicker  photometer,  the  comparison  standard  being  a  white  matt 
surface  (magnesia)  illuminated  by  means  of  a  tungsten  lamp. 
TABLE  III.— Relative  Brightness  of  Various  Papers 

IlvIyUMINATED   FROM   DIFFERENT  SOURCES. 

Color  of  paper 

White  Red  Yellow  Green  Blue 

Illuminant                                                  $  $  CJ>            <p            f 

Tungsten  (Ruby)  lamp 100  84  96        23          9 

Mercury-vapor  lamp 100          8  95         33         14 

Tungsten  lamp 100  36  91         26        11 

Carbon,  4  watts  per  candle  lamp 100  41  94         27         11 

The  papers  were  all  viewed  at  the  same  angle  and  likewise 
illuminated  from  the  same  direction.  Specular  reflection  was 
carefully  avoided.  The  results  for  the  carbon  and  tungsten  lamps 
are  the  averages  of  a  great  many  determinations. 

SUMMARY. 

Based  upon  purely  theoretical  considerations,  computations  are 
made  showing  the  influence  of  colored  surroundings  on  the  color 
of  the  useful  light.  The  color  of  the  multi-reflected  light  is  shown 
with  and  without  being  diluted  by  direct  light.  The  intensity  and 
color  value  of  the  light  is  calculated  for  various  reflections.  They 
were  roughly  verified  by  actual  measurements  of  the  color 
changes  of  light  in  a  large  box,  the  interior  of  which  was  covered 
with  various  combinations  of  white,  yellow,  and  green  papers. 
The  results  obtained  were  consistent  with  the  previous  com- 
putations. 

The  actual  change  in  the  reflection  coefficients  of  various  col- 
ored papers  when  illuminated  by  various  commercial  light  sources 
is  shown. 

The  writer  is  indebted  to  Mr.  Leonard  Krill  for  assistance  in 
the  experimental  work. 


LUCKIESH  I      INFLUENCE  OE  COLORED   SURROUNDINGS  "Jl 

DISCUSSION. 

Dr.  Herbert  E.  Ives  (communicated)  :  The  paper  by  Mr. 
Luckiesh  on  "The  Influence  of  Colored  Surroundings"  is  not,  I 
think,  as  clear  as  it  should  be  on  several  points  in  color  measure- 
ment. 

The  first  part  of  the  paper  contains  what  are  called  "purely 
theoretical  computations"  on  color  changes  to  be  expected  by 
multiple  reflection  of  light  from  colored  surfaces.  As  a  basis 
for  these  are  taken  certain  "colorimeter  readings."  As  a  matter 
of  fact  the  computations  are  not  for  a  colorimeter  at  all,  but  for  a 
spectrophotometer  making  measurements  at  three  wave-lengths. 
Instead  of  calling  these  "colorimeter  readings"  they  should  be 
called  "intensities  on  an  arbitrary  scale." 

Taking  up  this  matter  in  detail  one  finds  on  the  second  page 
that  a  special  illuminant  is  assumed,  radiating  three  mono- 
chromatic radiations — red,  green  and  blue.  On  the  third  page 
the  definition  of  the  quantities  NRQ'R,  etc.,  which  follows 
the  calculations  shows  that  the  "color  values"  are  supposed  to  be 
obtained  by  the  colorimeter  under  these  three  monochromatic 
lights.  Another  limitation  which  is  tacitly  made  is  that  the 
special  tri-chromatic  light  shall  be  composed  of  exactly  the  wave- 
lengths used  in  the  colorimeter. 

The  complete  assumption  is  then  for  a  tri-color  illuminant  of 
the  same  constituents  as  the  colorimeter  primaries,  and  that  the 
colorimeter  values  of  the  colored  surfaces  are  to  be  obtained 
under  this  illuminant.  This  means  nothing  more  nor  less  than 
that  the  colorimeter  has  been  ingeniously  transformed  into  a 
photometer  for  making  measurements  under  the  three  special 
colors  of  light — in  other  words,  into  a  form  of  spectrophotometer. 

The  essential  characteristic  of  a  colorimeter,  as  commonly  un- 
derstood, is  that  it  measures  colors  of  any  type  of  spectral  com- 
position in  terms  of  the  mixing  proportions  of  three  primaries,  not 
in  terms  of  the  relative  intensities  of  the  three  primary  wave- 
lengths, as  present  in  the  light.  A  colorimeter  will  measure  a 
pure  yellow  as  so  many  red  and  so  many  green,  where  a  spectro- 
photometer or  Mr.  Luckiesh's  assumed  instrument  would  find  no 
red  and  no  green.  This  illustration  will  show  the  caution  which 
must  be  observed  against  confusing  the  three-color  measurements, 


72  TRANSACTIONS  I.  D.  S. — PART  II 

on  which  calculations  have  been  made  in  this  paper,  with  colori- 
meter measurements  as  ordinarily  carried  out.  One  special 
limited  case  has  been  chosen  in  which  the  colorimeter  can  be  used 
as  a  spectrophotometer.  The  calculations  show  the  effect  of 
multiple  reflections  on  monochromatic  light.  A  three-fold  repeti- 
tion of  the  expression  representing  this  effect  with  the  introduc- 
tion of  the  subscripts  R,  G  and  B  does  not  make  it  apply  to  a 
colorimeter.  The  word  "colorimeter"  had  much  better  have 
been  left  out  of  this  part  of  the  paper. 

The  problem  which  Mr.  Luckiesh  has  studied  can  be  solved  by 
first  working  out  the  effect  through  the  whole  spectrum  (using 
many  more  than  three  wave-lengths),  plotting  the  results  as 
spectrophotometric  curves  and  then  determining  the  colorimeter 
values  either  through  the  color  sensation  curves,  or  by  the  color 
mixture  curves  of  the  colorimeter  used.  This  process  has  been 
published  in  some  detail  in  the  Transactions  of  the  Society. 

The  second  part  stands  alone  as  a  set  of  measurements  made 
with  a  different  instrument  than  that  "theoretically"  discussed  and 
only  related  to  the  first  part  in  a  very  general  manner  indeed. 
They  are  "consistent  with  the  theoretical  computations"  only  in 
that  qualitative  way  which  they  must  be  from  the  most  elementary 
considerations  of  the  phenomena  of  light  absorption. 

A  criticism  of  the  second  part  is  that  the  manner  of  use  of  the 
colorimeter  is  not  described.  The  essential  fact  to  know  about 
a  color  mixing  instrument  is  what  its  primary  mixing  colors  are. 
In  the  Ives  colorimeter  it  makes  a  difference  whether  the 
primaries  are  derived  from  daylight  or  from  a  tungsten  lamp,  or 
other  artificial  source.  The  description  of  the  instrument  as 
used  should,  therefore,  be  included  in  a  revision  of  the  paper. 

A  further  criticism  may  be  made  on  the  manner  of  plotting  the 
colorimeter  results.  They  are  shown  in  a  color  triangle  whose 
"white"  is  the  yellow  of  the  tungsten  lamp.  This  means  a 
triangle  not  heretofore  employed,  and  one  in  which  the  relation- 
ships are  much  distorted  from  those  in  the  usual  white  light 
triangle.  It  is  the  object  of  scientific  study  to  reduce  the  number 
of  elementary  quantities  to  the  lowest  possible,  and  to  establish 
exact  relationships  between  them.  The  employment  of  this  new 
form  of  triangle  without  establishing  the  relationship  of  its  read- 


LUCKIESH  :       INFLUENCE  OF   COLORED   SURROUNDINGS  /$ 

ings  to  that  of  a  white  light  triangle  is  to  be  regretted.  It  is  one 
thing  to  use  different  units  when  the  law  of  relationship  is  known ; 
it  is  quite  a  different  thing  when  the  law  is  not  known.  The 
relative  positions  of  points  on  a  straight  line  may  be  given  with- 
out stating  the  units  of  measurements,  because  the  relation  re- 
mains unchanged  no  matter  what  the  units.  In  going  from  one 
color  triangle  to  another  with  different  vertices  or  center,  the 
units  and  so  the  relative  positions  of  colors  suffer  distortions  not 
represented  by  any  simple  law.  It  should  therefore  not  be  over- 
looked that  measurements  plotted  in  this  triangle  are  not  directly 
comparable  with  those  plotted  in  the  usual  forms  of  color  triangle. 
The  measurements  are,  as  they  stand  much  more  qualitative  than 
quantitative. 

Mr.  M.  Luckiesh  (in  reply)  :  I  am  thankful  to  Dr.  Ives  for 
his  interest  in  the  paper. 

The  theoretical  computations  made  in  the  first  part  of  the 
paper  are  not  for  a  colorimeter  at  all  nor  do  I  make  such  a  state- 
ment. The  quantities  NRQ'R,  etc. ,  on  the  third  page  are  not  meant 
to  represent  any  colorimeter  measurements  of  mine.  They  do 
represent  the  quantity  usually  measured  in  determining  the  color 
of  paper.  I  am  well  aware  of  what  the  computations  represent, 
for  the  method  was  adopted  after  serious  consideration.  A  spe- 
cial illuminant  was  assumed  because  of  the  ease  in  illustrating 
just  what  I  desired  to  show  which  was  not  only  the  color  changes 
but  also  the  influence  of  multiple  reflections.  I  fully  appreciate 
there  are  other  methods  for  showing  these  effects,  but  I  chose 
this  one  as  being  the  most  desirable  from  my  viewpoint. 

Dr.  Ives'  discussion  contains  nothing  which  attacks  the  correct- 
ness of  my  computations.  No  claims  for  the  results  are  made 
knowingly  which  are  not  justified  by  the  assumptions. 

For  the  most  part  of  Dr.  Ives'  discussion  which  is  valuable  and 
scholarly,  consists  of  considerations  I  had  made  before  adopting 
the  method  but  which  I  deemed  as  unnecessary  to  embody  in  the 
paper. 

The  latter  part  of  Dr.  Ives'  discussion  takes  up  the  actual 
measurements  with  the  colorimeter.  I  am  not  in  a  position  as 
yet  to  make  my  colorimeter  measurements  accord  with  those  of 
Dr.   Ives.     Therefore  it  is   well  to  use  an   altogether  different 


74  TRANSACTIONS  I.  £.  S. — PART  II 

triangle.  It  is  true  the  measurements  plotted  on  this  triangle  do 
not  have  quite  the  same  spacial  relation  to  each  other  as  they 
would  have  on  the  triangle  used  by  Ives.  But  the  relative  posi- 
tions of  the  color  values  of  tungsten  and  carbon  incandescent 
lamps  on  my  triangle  are  sufficient  to  give  an  idea  of  the  order  of 
magnitude  of  the  changes  possible  due  to  colored  surroundings. 
If  the  other  points  are  shifted  more  in  one  triangle  than  in  the 
other  all  that  is  necessary  in  one  case  is  to  describe  the  green 
paper  as  a  greener  green  or  the  yellow  paper  as  a  yellower  yellow. 
It  seems  like  a  waste  of  time  to  quibble  over  that  point  consider- 
ing the  difficulties  of  accurately  describing  the  papers  used. 

The  methods  adopted  in  this  paper  are  legitimate  as  long  as  it 
is  plain  that  the  conclusions  are  limited  by  the  assumptions.  I 
intended  that  they  should  be  so  and  believe  they  are.  There  are 
more  ways  than  one  to  attack  this  problem  but  I  chose  a  method 
which  was  clear  cut  and  absolutely  free  from  speculation  and  am 
satisfied  that  it  best  illustrated  what  I  desired  to  bring'  forth. 


THOMAS  :      THEORY  OF  MERCURY-VAPOR  APPARATUS  75 

THEORY  OF  MERCURY- VAPOR  APPARATUS.* 


BY  PERCY   H.  THOMAS. 


Synopsis:  This  paper  gives  a  working  hypothesis  or  conception  of 
the  actions  going  on  in  a  mercury-vapor  apparatus,  from  the  point  of 
view  of  the  electron  theory  of  electricity.  According  to  the  hypothesis 
of  the  paper,  current  consists  of  electrons  passing  in  the  circuit  of  the 
apparatus  and  through  the  vacuum  from  the  cathode  to  the  anode  in  the 
form  of  material  particles.  The  resistance  in  the  vacuum  is  due  to  the 
obstruction  of  the  molecules  of  the  gas  or  vapor  in  the  space,  and  the 
pressure  of  the  vapor  is  the  ordinary  pressure  of  saturated  vapor  in  the 
presence  of  its  liquid,  as  the  mercury  electrode,  and  is  determined  solely 
by  the  temperature  of  the  mercury.  The  paper  also  explains  the  funda- 
mental characteristics  of  the  high  pressure  mercury-vapor  lamps  in  quartz 
tubes. 

The  theory  of  the  mode  of  operation  of  the  well  known  mer- 
cury-vapor apparatus,  characterized  by  a  hermetically  sealed  con- 
tainer exhausted  to  a  high  degree  of  purity  and  enclosing  suitable 
positive  and  negative  electrodes,  is,  more  than  that  of  most 
physical  apparatus,  dependent  upon  the  aid  of  the  electron 
hypothesis  of  the  nature  of  electricity.  The  characteristic  ob- 
served properties  of  this  apparatus,  as  distinguished  from  the 
theory  of  its  operation,  are  very  simple  in  the  fundamental  em- 
bodiment and  are  well  known. 

As  to  the  voltage  consumed  in  a  mercury-vapor  device,  it  may 
be  stated  that  the  drop  of  potential  is  practically  constant  for 
values  of  current  above  a  certain  minimum,  regardless  of  the 
strength  of  the  current,  provided  the  vapor  pressure  of  the 
mercury  vapor  is  maintained  constant.  Of  this  constant  voltage 
consumed  in  the  device,  there  is  a  certain  portion,  also  constant, 
consumed  at  each  electrode,  while  the  remaining  constant  portion 
is  consumed  in  the  passage  of  current  through  the  vacuum  or 
vapor  space.  Only  this  latter  portion  of  the  total  voltage  drop  is 
dependent  upon  the  length  and  diameter  of  the  vapor  path. 

The  above  statement  as  to  voltage  drop  applies  to  the  operating 
lamp.  Before  starting,  the  conditions  .are,  however,  entirely 
different.  •  The  device  operates  as  though  it  contained  somewhere 
a  rigid  obstruction  to  the  flow  of  current,  which  obstruction  sub- 

*  A  paper  read  at  a  meeting  of  the  New  England  section  of  the  Illuminating  Engi- 
neering Society.  Boston,  February  17,  1913. 


j6  TRANSACTIONS  I.  E.  S. — PART  II 

stantially  disappears  when  once  overcome.  A  study  of  the  be- 
havior of  this  apparatus,  particularly  its  behavior  when  utilized 
as  a  rectifier,  points  conclusively  to  the  surface  of  the  electrode 
impressed  with  potential  in  the  negative  direction,  that  is,  to  the 
cathode,  as  the  location  of  this  peculiar  obstruction  to  the  starting. 
This  fact  is  established  partly  by  the  observed  condition  that 
the  starting  obstruction,  or  "reluctance,"  as  it  is  called  by  Dr. 
Hewitt,  can  be  overcome  by  various  operations  at  the  surface  of 
the  cathode  and  can  not  be  overcome  by  any  operations  in  any 
other  part  of  the  device,  and  is  partly  demonstrated  by  the  fact 
that,  with  the  device  connected  according  to  the  ordinary  method 
as  a  rectifier,  namely,  with  two  anodes  connected  to  the  terminals 
of  the  supply  and  the  cathode  connected  to  an  intermediate  point 
of  the  supply,  current  will  flow  freely  in  any  direction  in  the 
vacuum  space  and  will  flow  out  of  any  electrode  impressed  with  a 
positive  potential  and  yet  current  will  not  flow  between  the  two 
anodes,  one  or  the  other  of  which  is  always  impressed  with  a 
positive  potential  while  the  other  has  a  negative  potential.  Since 
now  it  is  known  that  the  anode  impressed  with  the  positive  po- 
tential will  not  oppose  the  flow  of  current  from  the  anode  into 
the  vapor  and  since  it  is  known  that  current  can  flow  in  any 
direction  through  the  vacuum  space,  it  must  follow  that  the 
reason  the  rectifier  does  not  short  circuit  or  "arc"  between  the 
anodes,  in  other  words  the  reason  that  current  does  not  flow 
directly  between  the  main  anodes,  is  the  existence  of  some  sort 
of  obstruction  or  reluctance  at  the  anode  impressed  with  the 
negative  potential,  which  is  in  fact  the  case. 

One  other  characteristic  property  of  the  apparatus  I  wish  to 
bring  out,  namely,  that  the  voltage  consumed  in  the  vapor  path 
proper  depends  upon  the  vapor  pressure  or  density  of  the  mer- 
cury vapor  inside  the  container.  Since  there  is  liquid  mercury 
in  the  enclosed  space  and  no  gas  or  vapor  except  the  vapor  of 
mercury,  the  pressure  of  the  mercury  vapor  within  must  always 
be  the  pressure  corresponding  to  the  vapor  tension  of  mercury  at 
the  temperature  of  the  liquid  mercury  itself.  This  is  seen  to  be 
true  since,  were  the  vapor  pressure  less  than  the  appropriate 
value,  the  liquid  mercury  would  evaporate  until  the  pressure  of 
saturation  be  reached  and  were  the  pressure  in  the  vapor  greater, 
vapor  would  be  condensed  until  again  the  pressure  of  saturation 


THOMAS:       THEORY  OF  MERCURY-VAPOR  APPARATUS  J  J 

corresponding  to  the  temperature  of  the  electrode  be  obtained. 
Therefore,  the  only  way  to  increase  or  decrease  the  pressure  in 
a  mercury-vapor  tube  is  to  increase  or  decrease  the  temperature 
of  the  liquid  mercury. 

One  more  point.  When  mercury  evaporates  it  absorbs  heat ; 
when  it  condenses  it  liberates  heat,  as  in  the  case  of  any  liquid 
in  the  pressure  of  its  vapor;  consequently,  if  heat  be  generated 
in  liquid  mercury  within  the  condenser,  an  equivalent  amount  of 
heat  will  be  transferred  to  the  coolest  part  of  the  wall  of  the 
container  by  the  evaporization  of  mercury  at  the  electrode  and 
the  condensation  of  mercury  on  the  wall  of  the  container.  From 
this  it  follows  that  where  two  bodies  of  mercury,  as  for  example 
two  mercury  electrodes,  exist  in  the  same  device,  they  must 
necessarily  have  their  surfaces  at  approximately  the  same  tem- 
perature, since  otherwise  mercury  would  evaporate  from  the  hot 
electrode,  thus  cooling  it,  and  condense  on  the  cold  electrode, 
thus  heating  it.  Many  of  the  features  and  characteristics  of 
mercury-vapor  apparatus  can  be  explained  or  understood  by  a 
knowledge  of  these  principles.  For  example,  the  use  of  a  con- 
densing chamber  as  a  means  of  controlling  the  temperature  of 
the  lamp  that  is,  that  of  the  mercury  cathode,  the  vapor  pressure 
and  the  voltage  consumed  in  the  vapor  path,  may  be  clearly 
understood. 

It  is  now  in  order  to  consider  the  electrical  action  of  the 
operation  of  the  device,  which  for  the  present  discussion  may  be 
assumed  to  be  a  lamp.  A  flow  of  electricity,  according  to  the 
electron  hypothesis,  to  which  I  personally  subscribe,  is  nothing 
more  than  the  passage  of  electrons  along  a  circuit.  It  is  known 
that  these  electrons  are  physical  bodies  of  extremely  small  mass, 
requiring  between  one  and  two  thousand  to  give  the  mass  of 
the  hydrogen  atom,  and  carry  a  definite  negative  electrical  charge. 
These  electrons  exist  in  a  relatively  quiescent  state  in  all  matter. 
Since  the  charge  carried  by  the  electrons  is  negative  and  since 
unelectrified  bodies  show  no  resultant  electric  charge,  it  must 
follow  that  in  such  unelectrified  matter-  the  negative  charge  on 
the  electron  is  balanced  by  some  positive  charge  in  the  material. 

There  will  be  no  flow  of  electricity  in  an  electric 
circuit,  that  is,  no  flow  of  electrons,  until  something  disturbs  the 
condition  of  electrical  balance  of  the  unelectrified  matter.  When- 


78  TRANSACTIONS  I.  E.  S. — PART  II 

ever,  however,  the  electron  is  separated  from  the  positive  charge 
which  ordinarily  neutralizes  its  external  electrical  effect  and 
when  there  exists  at  the  same  time  an  electro-motive-force  in  the 
neighborhood  of  this  electron,  the  electron  will  endeavor  to  fol- 
low the  electro-motive-force  and  produce  a  flow  of  current.  But 
electrons  like  other  material  particles  or  bodies  can  not  move  un- 
less they  have  a  free  space  in  which  to  move.  In  all  metal  con- 
ductors it  appears  that  there  are  passages  or  spaces  between  the 
atoms  or  molecules  through  which  the  electrons  can  pass  re- 
latively freely,  although  they  will  experience  some  resistance  (the 
well  known  ohmic  resistance  of  the  metals).  Electrons  set  free 
in  insulating  material,  however,  do  not  find  any  passageways 
open  for  motion  from  one  place  to  another,  hence  the  electrons 
with  their  charges  remain  fixed  in  location.  This  is  illustrated, 
for  example,  by  the  rod  of  sealing  wax  and  the  catskin.  When 
rubbed  by  the  skin  the  rod  becomes  electrified,  but  the  electricity 
remains  fixed  in  position.  Under  this  condition  the  electrons  are 
stuck,  so  to  speak,  in  the  insulating  material. 

Although  air  is  ordinarily  an  insulator  it  is  generally  assumed 
that  air  and  other  gases  may  be  made  under  certain  conditions 
to  be  good  conductors  of  electricity  and  while,  in  a  certain  sense, 
that  is  true,  it  is,  nevertheless,  not  analytically  true.  The  prin- 
cipal reason  that  electricity  does  not  flow  through  ordinary  air 
and  gases  is  that  there  are  no  free  electrons  in  the  gas  to  move  in 
response  to  an  electro-motive-force.  When  free  electrons  are" 
introduced  or  produced  in  a  gas,  they  do  pass  along  through  the 
gas  if  a  suitable  electro-motive-force  is  present.  They  pass,  how- 
ever, not  through  the  atoms  themselves  but  in  the  space  between 
them.  This  action  is  very  much  complicated,  however,  by  the 
fact  that  part  of  the  progress  may  be  due  to  the  movement  of 
the  molecules  of  the  air  itself  and  by  the  fact  that  electrons  have 
an  attraction  for  gas  molecules  and  stick  to  them,  sometimes  col- 
lecting quite  a  group  of  molecules  called  an  aggregate.  The  trans- 
fer of  electricity  in  the  form  of  electrons,  freed  or  liberated  in 
air,  is  illustrated  by  the  progress  of  a  thunder  cloud  where  the 
electric  charges  travel  considerable  distances  with  the  air  and 
probably  to  some  extent  through  the  air  itself,  slipping  between 
the  molecules.     If  electrons  be  set  free  from  the  atoms  of  the 


THOMAS:      THEORY  OF  MERCURY-VAPOR  APPARATUS  /9 

gas,  as  can  be  done  with  the  aid  of  X-rays,  and  an  electro- 
motive-force be  applied,  as  was  done  by  J.  J.  Thomson  in  his 
famous  experiments  in  which  he  applied  electro-motive-forces 
of  opposite  signs  to  parallel  plates,  there  will  be  an  actual  flow 
of  current  through  the  air  due  to  the  passage  of  electrons  through 
the  air.  Such  currents  are  always  very  small,  however,  since  up  to 
the  present  time  no  method  has  been  devised  for  producing  large 
quantities  of  free  electrons  from  gases,  under  any  such  circum- 
stances. 

In  the  electric  arc  in  air  a  large  number  of  electrons  are 
liberated  from  the  cathode  and  force  a  lane  or  passageway  to  the 
anode,  probably  by  crowding  back  the  molecules  of  air. 

The  electron  and  its  neutralysing  positive  charge,  when  in  a 
state  of  equilibrium  and  constituting  a  state  of  non-electrification, 
have  an  attraction  for  each  other  and  can  not  be  separated  with- 
out the  exertion  of  a  considerable  force  and  the  expenditure 
of  a  certain  amount  of  energy  and  ionization,  as  such  sepa- 
ration is  called,  is  ordinarily  a  difficult  process.  The  forces 
being  inter-  or  infra-automatic  are  very  large  in  propor- 
tion to  the  physical  size  of  the  electrons.  Furthermore, 
it  is  difficult  to  employ  the  powerful  forces  available  in 
connection  with  large  masses  of  matter  in  such  a  way  as  to  be 
effective  in  separating  the  electrons  from  an  atom.  Electro- 
magnetic waves  of  very  short  wave-lengths  seem  to  be  effective 
in  producing  this  result.  Ionization  may  be  produced  by  such 
waves  as  are  supplied  by  X-ray  apparatus  or  by  ultra-violet  light. 
Another  very  effective  method  of  producing  ionization,  that  is 
the  separation  of  electrons  from  atoms,  is  the  shock  caused  by 
the  striking  of  one  atom  by  another  atom  or  by  an  electron  pro- 
ceeding at  a  very  high  velocity. 

Electrons  are  always  free  and  able  to  move  within  metallic  con- 
ductors as  long  as  they  do  not  go  outside. 

Now  the  electric  circuit  containing  a  vapor  electric  device 
having  a  high  vacuum  between  two  electrodes  may  be  considered. 
If  an  electro-motive-force  be  impressed  in  such  a  circuit  the  elect- 
rons will  flow  freely  in  the  metal  parts  of  the  circuit 
from  a  point  of  low  potential  to  the  point  of  high 
potential    (they    move    backward    on    account    of    their    carry- 


8o  TRANSACTIONS  I.  £.  S. — PART  II 

ing  a  negative  charge)  until  they  reach  some  point  where 
their  progress  is  blocked  that  is  where  the  electrons  are  not  free 
to  move.  They  are  free  to  move,  however,  in  all  that  part  of 
the  circuit  constituted  by  metal  conductors.  Following  the 
electrons  within  these  metal  conductors  one  finds  them  flowing 
freely  toward  the  cathode  of  the  device  until  they  reach  the  sur- 
face of  the  cathode  exposed  in  the  vacuum.  But  the  electron  can 
not  leave  the  surface  of  the  cathode  without  overcoming  the  at- 
traction of  this  electron  for  the  corresponding  positive  charge 
associated  with  it.  It  could  move  freely  in  the  body  of  the  metal 
since  in  leaving  one  positive  charge,  it  could  pick  up  another 
from  an  adjacent  atom.  These  electrons  then  accumulate  at 
the  cathode  surface  producing  there  a  negative  charge;  similarly 
electrons  are  withdrawn  from  the  anode  producing  there  a  posi- 
tive charge,  which  two  charges  impress  the  electro-motive- force 
of  the  circuit  on  the  vapor  path  in  the  vacuum  space.  Now  were 
there  a  supply  of  free  electron  in  the  vacuum  space,  they  would 
be  immediately  drawn  out  of  the  vacuum  space  into  the  anode 
producing  a  flow  of  current  as  long  as  the  supply  lasted.  As  has 
already  been  pointed  out  there  is  no  material  opposition  or  resist- 
ance to  the  entering  of  a  metal  conductor  by  a  free  electron  since 
no  counter  attraction  for  a  positive  charge  must  then  be  over- 
come. If,  however,  the  electrons  which  have  accumulated  at  the 
surface  of  the  cathode  could  overcome  the  attraction  they  have 
for  their  positive  charges  and  get  into  the  vacuum  space  they 
would  leave  the  cathode  and  there  would  immediately  be  a  stream 
of  electrons  between  the  electrodes  in  the  vacuum  space  giving  a 
flow  of  current  and  there  would  be  no  limit  to  the  amount  of 
this  current  unless  a  limit  developed  in  the  supply  of  electrons. 
But  the  supply  of  electrons  is  unlimited  in  the  metals. 

Now  when  a  mercury-vapor  lamp  is  started  into  operation, 
this  means  merely  that  a  means  has  been  provided  for  liberating 
electrons  from  the  surface  of  the  cathode.  Then  these  electrons 
are  free  to  flow  under  the  influence  of  the  electro-motive-force  of 
the  circuit  from  the  cathode  through  the  vapor  path  to  the  anode 
and  into  the  anode  through  the  metallic  circuit  outside,  back 
through  the  cathode  lead  to  the  cathode  again.  In  this  circuit 
there  is  developed,  of  course,  a  certain  amount  of  resistance  in 


THOMAS:       THEORY  OF  MERCURY- VAPOR  APPARATUS  8l 

each  part  of  the  circuit;  the  well  known  ohmic  resistance  in  the 
metallic  conductors ;  a  certain  resistance  to  the  liberation  of 
electrons  at  the  cathode  surface,  and  a  certain  resistance  to  the 
passage  of  electrons  through  the  vapor  space.  This  latter  resist- 
ance, namely  the  resistance  to  the  passage  of  electrons  (or  cur- 
rent) through  the  vapor  space  results  principally  from  the 
jostling  and  blocking  of  the  electrons  by  the  atoms  or  molecules 
of  the  vapor  present  which  get  in  the  path  of  the  electrons. 
The  lower  the  pressure  of  the  vapor  the  less  this  resistance  and 
the  less  the  voltage  absorbed  in  the  vapor  path;  the  higher  the 
vapor  pressure,  the  greater  the  tendency  to  impede  the  progress 
of  the  electrons  and  the  resistance  of  the  lamp  or  its  voltage  drop. 

In  a  lamp,  however,  it  is  this  jostling  of  the  vapor  molecules 
that  produces  the  light  and  the  more  vigorously  they  are  jostled, 
that  is,  the  greater  the  volume  of  the  current  flow  and  the  greater 
the  number  of  electrons,  the  greater  the  amount  of  light ;  and 
again  the  greater  the  number  of  vapor  molecules,  that  is  the 
greater  the  vapor  pressure,  the  greater  the  amount  of  light. 

After  starting  up  a  mercury-vapor  lamp  cold,  although  there 
is  at  first  an  abnormally  large  current,  there  is  very  little  light 
produced.  This  is  because  there  is  very  little  vapor  present  and 
a  relatively  small  number  of  vapor  molecules  are  jostled.  As 
the  lamp  warm?  up,  however,  although  the  current  becomes 
somewhat  less,  the  amount  of  light  given  is  far  greater  since  the 
number  of  molecules  of  vapor  is  greatly  increased  on  account 
of  the  higher  vapor  pressure. 

The  warming  up  process  comes  to  a  stage  of  equilibrium  when 
the  heat  radiated  or  dissipated  from  the  surface  of  the  lamp 
equals  the  heat  generated  in  the  lamp.  If  now  the  heat  dissi- 
pating capacity  (for  example  by  the  use  of  a  condensing  cham- 
ber) of  the  device  is  so  proportioned  that  this  equilibrium  is 
reached  when  the  mercury  temperature  is  somewhat  above  the 
boiling  point  of  water  one  has  a  mercury-vapor  lamp  of  the  low 
pressure  type;  if  on  the  other  hand  the  heat  dissipating  power 
of  the  lamp  is  reduced  so  that  equilibrium  is  reached  at  a  con- 
siderably higher  temperature,  there  is  obtained  the  high  pressure 
type  of  mercury-vapor  lamp,  the  type  for  which  a  quartz  con- 
tainer may  be  advantageously  used. 


82  TRANSACTIONS  I.  E.  S. — PART  II 

Returning  now  to  the  surface  of  the  negative  electrode  and 
the  means  by  which  the  starting  reluctance  of  the  cathode  is 
overcome  and  electrons  are  freed  from  the  body  of  the  cathode 
material,  it  is  necessary  to  confess  that  the  exact  mechanism  of 
this  process  is  not  known  with  certainty.  A  prominent  char- 
acteristic feature  of  the  process  is,  however,  the  so-called  cathode 
spot  or  bright  spot  of  light  at  the  point  where  the  electrons  leave 
the  cathode  surface,  which  is  one  of  the  features  distinguishing 
this  light  from  the  so-called  Geisler  tubes.  At  this  spot  some- 
thing is  going  on  which  is  liberating  electrons  from  their  asso- 
ciated positive  charges  in  the  atoms  of  the  liquid  mercury.  It 
may  be  that  the  heat  generated  by  the  current  flow  concentrated 
at  this  point  produces  a  very  dense  vapor  and  that  the  current 
which  is  greatly  concentrated  at  this  point  serves  to  ionize  this 
concentrated  mercury  vapor  very  energetically,  this  liberating 
of  electrons  serving  to  secure  the  continued  flow  of  current  in 
the  vapor  space.  In  any  event  there  is  some  result  of  the  flow 
of  current  at  any  one  instant  which  provides  for  the  liberation 
of  electrons  to  constitute  the  flow  of  current  during  the  next 
instant.  Whether  this  be  an  extremely  local  heat  effect  or  the 
rapid  ionizing  of  vapor  generated  locally  or  whether  it  be  the 
liberation  of  electrons  directly  from  the  liquid  mercury  by  the 
bombardment  of  other  electrons  or  positively  charged  atoms  has 
not  been  determined.  It  is  interesting  to  remember,  however, 
that  if  there  is  an  extremely  plentiful  ionization  of  vapor  at  the 
cathode  spot  there  will  be  produced  by  the  current  flow  first 
electrons  which  will  be  attracted  to  the  anode  and  second  there 
will  be  liberated  by  these  electrons  the  corresponding  positively 
charged  atoms,  which  will  be  attracted  to  the  cathode  by  its 
negative  charge.  It  may  be  that  these  latter  atoms  which  must 
be  continuously  bombarding  the  cathode  are  the  means  of  liber- 
ating electrons  from  the  cathode  to  support  the  flow  of  current. 
However  this  may  be,  the  fact  can  hardly  be  controverted  that 
the  essential  action  which  eliminates  the  initial  starting  reluctance 
is  closely  related  to  some  mechanism  operating  in  the  cathode 
spot  and  self  perpetuating  when  once  started,  as  long  as  a 
flow  of  current  in  sufficient  volume  continues. 

It  is  a  well  known  fact  that  if  an  attempt  is  made  to  start  a 
flow  of  electricity  through  an  extremely  highly  exhausted  space 


THOMAS:      THEORY  OF  MERCURY-VAPOR  APPARATUS  83 

that  enormous  potentials  are  required.  It  was  customary  orig- 
inally to  attribute  this  phenomenon  to  the  supposed  absence  of 
a  conductor  in  the  vacuum  space,  but  our  present  electron 
hypothesis  has  shown  that  electricity,  that  is  electrons,  being 
physical  bodies  move  with  the  greatest  facility  in  a  vacuum  and 
that  the  reason  that  the  vacuum  device  resists  the  initial  flow  of 
current  so  stubbornly  is  the  fact  that  no  means  exist  for  lib- 
erating electrons  within  the  vacuum  space  from  the  surface  of 
the  cathode  which  is  the  only  point  at  which  they  can  be  pro- 
duced, since  there  is  no  gas  or  vapor  in  the  vacuum  which  can  be 
ionized  to  produce  electrons.  The  high  voltage  required  for  start- 
ing in  the  high  vacuum  is  to  be  expected  since  it  is  only  by  forces 
acting  directly  on  the  atoms  themselves  that  electrons  can  be  pro- 
duced from  solid  or  liquid  materials  and  a  very  high  starting 
voltage  must  be  provided  since  it  must  be  applied  at  a  distance. 

When,  however,  the  vacuum  in  the  device  just  discussed  is 
not  perfect  and  a  certain  residual  gas  is  present,  the  high  voltage 
applied  to  the  terminals  is  sufficient  to  ionize  this  gas  producing 
electrons  and  positively  charged  atoms.  These  positively  charged 
atoms,  as  already  described,  will  be  attracted  to  the  cathode  sur- 
face where  they  will  bombard  the  material  of  the  cathode,  thus 
liberating  further  electrons  and  positive  charges  which  repeat 
the  process  until  under  favorable  conditions  the  permanent  con- 
dition of  current  flow  as  already  described  is  attained.  The 
nature  of  these  phenomena  explains  why  the  salient  starting 
characteristics  of  the  cathode  in  a  vacuum  are  not  observed  in 
electrodes  in  the  open  air,  at  any  rate  to  anything  like  the  same 
extent  as  in  the  vacuum.  The  presence  of  the  air  between 
electrodes  provides  a  source  of  electrons  and  positive  charges 
automatically  sufficient  to  liberate  electrons  from  the  cathode, 
whenever  a  suitable  voltage  is  applied.  Furthermore,  the  pres- 
ence of  the  molecules  of  air  in  the  path  of  the  current  when  once 
started  so  greatly  increase  the  operating  voltage  that  the  effect 
of  the  starting  reluctance  would  be  practically  overshadowed. 

With  this  exposition  of  the  hypothesis  or  conception  of  the 
nature  of  the  operation  of  a  mercury-vapor  device  which  has 
satisfied  me  personally  and  seems  consistent  with  practically  all 
the  fundamental  principles  now  established  in  electro  physics, 
as  far  as  I  know  them,  there  remains  very  little  to  be  said  in 
3 


84  TRANSACTIONS  I.  E.  S. — PART  II 

explanation  of  the  theory  of  operation  of  the  practical  mercury- 
vapor  lamp,  either  the  low  pressure  lamp  or  the  high  pressure 
lamp  in  the  quartz  container. 

It  may  be  well  to  call  attention,  however,  to  one  characteristic 
of  the  quartz  burner  of  great  practical  importance  in  its  opera- 
tion, though  purely  incidental  in  the  electrical  hypothesis  involv- 
ing its  principle  of  operation.  I  refer  to  the  fact  that  the  quartz 
burner  operating  as  intended  in  the  Cooper  Hewitt  commercial 
quartz  lamp,  is  approximately  a  constant  current  device.  That 
is  if  the  voltage  applied  in  such  a  lamp  is  raised  the  only  effect 
is  to  increase  the  voltage  on  the  tube  without  material  change 
of  current  through  the  tube.  Of  course,  the  first  momentary 
effect  of  the  increase  in  voltage  is  an  increase  in  the  current, 
but  this  increase  in  current  raises  the  temperature  of  the  lamp 
thus  increasing  the  temperature  of  the  mercury  electrode  and 
the  pressure  of  the  mercury  vapor.  This,  in  turn,  increases 
the  resistance  of  the  lamp  or  the  voltage  consumed  therein  and 
the  point  of  equilibrium  is  found  finally  at  a  current  only  slightly 
greater  than  the  original  current  flow.  Lowering  of  the  voltage 
merely  produces  a  lowered  voltage  on  the  tube  when  equilibrium 
is  finally  attained  with  but  a  slightly  decreased  current.  These 
same  characteristics  are  found  in  the  low  pressure  lamp  but  are 
there  not  as  marked.  The  particular  significance  of  this  phe- 
nomenon lies  in  the  fact  that  if  an  attempt  is  made  to  adjust 
the  series  resistance  of  a  high  pressure  lamp  which  is  being 
installed  to  the  proper  value  by  the  insertion  of  an  ammeter  in 
the  circuit,  it  will  be  impracticable  to  make  a  satisfactory  adjust- 
ment, since  the  difference  between  the  current  shown  on  the  am- 
meter with  a  very  low  value  of  the  series  resistance  and  that  with 
a  very  high  value  will  be  very  slight  indeed.  On  the  other  hand, 
if  a  voltmeter  be  placed  in  shunt  to  the  burner  it  is  possible  to 
adjust  the  series  resistance  with  great  accuracy  and  certainty, 
since  the  voltage  on  the  burner  is  very  sensitive  to  the  proper 
setting.  With  the  use  of  a  voltmeter  it  is  not  necessary  to  pay 
any  attention  to  the  current,  for  this  will  take  care  of  itself. 

If  it  is  for  any  reason  desired  to  increase  the  current  in  a 
high  pressure  burner  (that  is  the  tube)  this  must  be  done  by 
increasing  the  natural  dissipation  of  heat  from  the  burner  to 
give  it  a  lower  temperature  as  by  placing  it  in  a  cooler  place  or 


THOMAS  :      THEORY  OF  MERCURY-VAPOR  APPARATUS  85 

by  directing  a  draft  on  the  tube  or  otherwise.  In  such  a  case 
the  net  result  of  the  cooling  is  to  lower  the  temperature  of  the 
electrode  and  the  pressure  of  the  mercury  vapor.  The  result  of 
the  lowering  of  the  voltage  or  the  pressure  is  an  increase  in  the 
current.  This  increase  in  the  current  will  then  heat  up  the  burner 
until  the  vapor  pressure  and  the  burner  voltage  again  bears  the 
right  relation  to  the  supply  voltage.  Vice  versa,  when  placed  in 
an  abnormally  hot  atmosphere  the  burner  will  take  an  abnormally 
small  current.  Thus,  in  very  cold  weather  or  with  a  cracked 
globe  the  tendency  of  a  quartz  burner  is  to  take  a  large  amount 
of  current  while  its  voltage  may  remain  approximately  normal. 

Another  result  of  this  characteristic  of  the  quartz  burner  is  the 
difficulty  of  running  the  constant  potential  quartz  burners  in 
series;  for  suppose  a  number  of  such  burners  to  be  placed  in  a 
constant  current  circuit;  some  of  them  will  naturally  run  a  little 
hotter  than  others  either  from  variations  in  the  structure  or  from 
different  temperature  conditions  at  their  points  of  installation. 
Those  naturally  running  hotter  will  be  taking  a  little  too  much  cur- 
rent and  those  running  cooler  will  be  taking  too  little  current,  but 
all  are  forced  to  take  the  same  current  since  they  are  a  constant 
current  circuit.  Since  now  these  devices  are  constant  current 
devices,  naturally  those  taking  too  much  current  will  heat  more 
and  more  and  the  hotter  they  tend  to  get  (since  the  added  vapor 
pressure  and  resistance  from  the  added  temperature  increase  th& 
heat  generated  in  the  burner,  even  if  there  be  no  increase  of 
current).  It  thus  soon  happens  that  a  few  burners  take  nearly 
all  the  potential  and  perhaps  ultimately  get  so  hot  as  to  put  out 
the  whole  series.  The  same  general  difficulty  was  originally  met 
with  in  arc  lamps,  but  was  overcome  by  the  use  of  shunt  regu- 
lating coils  coupled  with  means  for  adjusting  the  length  of  the 
arc  between  electrodes.  The  constant  potential  type  of  mercury 
quartz  burner  does  not  provide,  however,  for  such  adjustment 
and  this  method  is  inapplicable. 

DISCUSSION. 

Mr.  C.  F.  LorEnz  (communicated)  :  It  is  interesting 
to  see  that  the  electron,  which  has  come  to  play  such  a 
familiar  role  in  the  every  day  thought  of  physicists,  is  also  com- 
mencing to  penetrate  into  the  transactions  of  technical  societies. 


86  TRANSACTIONS  I.  E.  S. PART  II 

The  author  presents  a  very  clear  and  simple  picture  of  the  inner 
nature  of  the  things  going  on  in  a  mercury-vapor  tube,  but  in 
doing  so  he  throws  an  even  greater  burden  on  the  electron  than 
is  customary.  He  speaks  of  the  flow  of  current  as  being  entirely 
constituted  by  the  motion  of  electrons.  Ordinarily  we  think  of 
both  positive  and  negative  carriers  as  entering  into  nearly  every 
kind  of  electric  discharge;  an  exception  is  the  cathode  ray  stream 
in  an  X-ray  tube;  another  is  the  discharge  of  a  negatively  charged 
incandescent  solid  when  located  in  a  very  high  vacuum.  The 
author  of  the  paper  recognizes  the  presence  of  positively  charged 
particles  in  the  mercury-tube  when  he  speaks  of  the  bombard- 
ment of  the  cathode  as  the  source  of  ionization  at  the  cathode 
during  the  arc-discharge ;  why  should  such  positive  charges  not 
take  part  in  the  process  of  conduction  throughout  the  tube? 

Thinking  of  positive  and  negative  charges  leads  to  an  explana- 
tion of  the  "reluctance"  which  differs  from  the  author's.  Before 
the  arc  is  started  there  must  be  ionization  to  some  extent  and  the 
application  of  the  difference  of  potential  causes  motion  of  the 
positive  and  negative  charges  which  constitutes  the  minute  cur- 
rent then  flowing;  a  resistance  must  immediately  develop  at  both 
electrodes  owing  to  the  scarcity  of  ions  which  immediately  exists 
in  the  regions  adjacent  to  the  electrodes,  since  the  solid  electrodes 
do  not  furnish  carriers  to  replace  those  swept  out  of  these  regions 
in  the  act  of  conduction.  The  cathode  drop  is  enormously  greater 
than  the  anode  drop  because  of  the  much  greater  mobility  of  the 
negative  carriers. 

A  point  discussed  in  the  paper  of  much  general  interest  apart 
from  its  importance  in  the  operation  of  the  quartz  mercury 
"burner"  is  the  behavior  of  the  latter  as  an  approximately  con- 
stant current  device.  Carbon  incandescent  lamps,  Nernst 
glowers,  metalized  filament  lamps,  and  metal  filament  lamps  all 
have  their  own  mode  of  behavior  under  varying  voltage,  which 
makes  them  useful  as  resistances  for  special  purposes,  at  least  in 
experimental  work.  The  very  peculiarly  behaving  quartz  mer- 
cury lamp  is  a  welcome  addition  to  the  list  of  automatically  vary- 
ing resistors. 

Dr.  E.  Weintraub  (communicated)  :  The  electronic  theory 
of  the  operation  of  the  mercury  arc  given  by  Mr.  Thomas  does 


THEORY    OF    MERCURY-VAPOR   APPARATUS  87 

not  materially  differ  from  the  one  current  at  present  among  those 
working  in  the  field  and  would,  therefore,  call  for  little  comment. 
I,  myself,  have  on  different  occasions  expressed  similar  views. 

It  is  true  that  the  electronic  picture  of  the  arc  is  somewhat 
vague  and  indefinite,  that  especially  the  role  of  the  positive 
electricity  is  not  clear,  but  these  imperfections  can  be  ascribed 
and  perhaps  justly  so  to  the  inherent  complexity  of  the  arc.  From 
the  point  of  view  of  the  engineer  and  inventor  a  more  serious 
objection  to  the  theory  is  the  little  help  it  has  offered  so  far  in 
the  technical  development  of  the  mercury  arc  and  other  arcs.  I 
for  one  had  to  work  my  own  way  by  direct  study  of  the 
phenomena  and  by  the  analogy  method  of  reasoning. 

However,  the  electronic  theory  is  the  theory  of  our  age,  is  a 
partial  expression  of  the  truth,  gives  a  new  cross-section  through 
the  infinite  complexity  of  the  natural  phenomena  and  whatever 
its  ultimate  fate  may  be  one  can  hardly  do  better  at  present  than 
to  accept  it  and  use  it  as  far  as  feasible. 

It  is,  however,  to  be  regretted  that  Mr.  Thomas  found  it  neces- 
sary to  obscure  his  electronic  exposition  by  still  retaining  the  con- 
ception of  a  "starting  reluctance"  at  the  cathode  which  has  no 
place  in  the  electronic  theory  of  the  arc  and  for  that  matter  in 
any  scientific  theory.  The  author  himself  later  explains  that  the 
evacuated  space  contains  no  conductive  matter,  that  the  latter  has 
to  be  produced  at  the  cathode  and  that  an  expenditure  of  energy 
is  therefore  necessary  at  or  near  the  cathode.  If  this  justifies 
the  assumption  of  a  starting  reluctance  then  the  whole  universe 
is  full  of  "reluctances."  Whenever  for  lack  of  energy  or  for 
lack  of  another  sufficient  reason  a  certain  transformation  does 
not  happen  then  there  would  be  an  equal  reason  to  postulate  a 
"reluctance." 

With  respect  to  series  operation  of  quartz  mercury  arc  lamps 
the  evil  of  running  away  on  constant  current  which  is  described 
by  Mr.  Thomas  is  less  pronounced  in  the  case  of  the  new  type  of 
quartz  lamp  that  I  have  developed  recently  and  what  is  left  of  it 
is  overcome  by  special  automatic  regulating  means  so  that  at 
present  the  quartz  mercury  arc  lamp  is  available  for  both 
alternating  current  and  direct  current  series  circuits. 


88  TRANSACTIONS  I.  E.  S. — PART  II 

STREET  LIGHTING  WITH  ORNAMENTAL  LUMINOUS 
ARC  LAMPS* 


BY  C.  A.   B.  HAIvVORSON,  JR. 


Synopsis: — This  paper  outlines  the  general  lighting  requirements  of 
business  and  residential  streets  and  parkways.  For  business  streets  it  is 
contended  (i)  that  the  illumination  should  be  of  a  different  color  and 
character  from  that  employed  in  store  windows  in  order  that  the  effect  of 
the  latter  may  not  be  impaired,  (2)  that  the  illumination  should  be  com- 
paratively brilliant,  though  not  greater  than  the  intensity  of  the  windows, 
to  attract  trade  and  insure  traffic  safety.  Residential  street  lighting,  in  the 
opinion  of  the  author,  requires  the  use  of  as  few  light  sources  as  possible  to 
produce  the  average  low  intensity  of  illumination,  and  that  a  non-uniform 
illumination  is  more  desirable  than  an  extremely  uniform  one.  Parkway 
and  driveway  lighting  demand  primarily  an  illumination  of  sufficient  in- 
tensity from  light  sources  of  low  intensic  brilliancy  to  insure  traffic  safety. 
In  all  three  classes  of  lighting  the  decorative  possibilities  of  ornamental 
lamps  and  standards  is  said  to  deserve  particular  consideration. 

In  planning  a  system  of  exterior  illumination  to  meet  the  re- 
quirements of  the  city  of  to-day,  the  needs  of  each  section 
of  the  city  must  be  carefully  considered.  The  shopping  centres 
demand  one  type  of  illumination  peculiarly  their  own ;  the  re- 
sidential streets,  open  parks,  drives  and  outlying  districts  each  in 
their  turn  require  special  consideration  and  quite  different  treat- 
ment, as  regards  both  the  illuminating  units  and  the  ornamental 
standards  or  fixtures  employed. 

Unfortunately,  it  is  quite  impossible  by  the  use  of  illuminating 
data  alone  to  show  exactly  what  system  of  lighting  should  be 
employed  in  each  case  to  accomplish  the  best  results,  as  frequently 
certain  psychological  and  physiological  requirements  play  parts 
of  vast  importance. 

The  lighting  of  business  streets  requires,  in  addition  to  the 
usual  purposes  of  good  general  lighting  (police  protection,  ability 
to  read  easily  and  distinguish  persons,  etc.)  an  illumination  toler- 
able and  pleasing  to  the  eye,  yet  sufficiently  brilliant  to  produce  a 
marked  effect  in  the  improvement  of  business — obviously  a  mat- 
ter of  great  general  interest  to  the  city — by  attracting  people  to 
the  brilliantly  lighted  thoroughfares  and  at  the  same  time  allow- 

*  A  paper  read  at  a  meeting  of  the  New  England  section  of  the  Illuminating  Engi- 
neering Society,  Boston,  February  17,  1913. 


halvorson:     street  lighting  89 

ing  their  attention  to  be  drawn  freely  to  the  matter  of  most  im- 
portance, which  is  of  course,  the  attractive  window  dis- 
plays and  decorations.  This  object  is  largely  attained  by 
means  of  the  contrast  between  the  color  schemes  employed  in 
these  decorations  and  the  color  of  the  general  illumination.  It 
follows,  therefore,  that  general  illumination  must  be  furnished 
that  differs  in  color  from  the  illumination  obtained  from 
the  small  lighting  units  usually  employed  in  the  mer- 
chant's local  display,  and  also  differing  from  and  not 
interfering  with  the  color  effects  obtained  by  reflection 
or  secondary  illumination  from  the  goods  displayed.  White  light 
falling  on  the  show  window  from  without  can  only  enhance  the 
color  values  by  showing  them  correctly.  Obviously,  then  the  color 
and  quality  of  the  light  play  a  most  important  part  in  successful 
lighting  of  this  kind;  accordingly  white  light  of  low  intrinsic 
brilliancy  must  be  employed  in  order  to  obtain  the  best  results. 

The  intensity  of  the  illumination,  perhaps,  should  be  con- 
sidered of  next  importance,  for  the  ratio  of  gen- 
eral illumination  to  window  illumination  should  be  such 
that  the  window  illumination  far  outweighs  the  gen- 
eral illumination.  As  good  window  display  lighting  re- 
quires in  the  neighborhood  of  15  foot-candles,  it  is  highly  im- 
probable that  any  economical  scheme  of  general  illumination  that 
could  be  obtained,  would  approach  this  figure.  A  possible  ex- 
ception is  yellow  flame  arc  lighting;  but  this,  on  account  of  the 
color,  would  be  highly  unsuitable,  assuming  that  such  illuminants 
were  placed  low,  as  they  usually  are  in  this  country,  and  in  the 
direct  range  of  vision ;  in  which  case  an  immense  volume  of  light 
would  be  directed  toward  the  show  windows. 

Other  important  considerations  are:  the  appearance  of  the 
lighted  unit;  the  illumination  of  the  building  fronts  from  both 
esthetic  and  economic  viewpoints,  the  latter  particularly  as  re- 
gards the  benefits  accruing  the  upper-floor  tenants;  the  daytime 
appearance  of  the  lighting  standard,  dignified,  simple,  or  ornate, 
as  harmony  with  its  surroundings  requires,  and  yet  unobtrusive 
and  free  from  overhanging  arms  and  glassware  that  might  impede 
teaming  and  endanger  pedestrians.  Above  all  the  lighting  unit 
must  aid  in  beautifying  the  street  rather  than  produce  the  effect 


90  TRANSACTIONS  I.  E.  S. — PART  II 

of  crowding  and  over-burdening  which  is  so  characteristic  of 
many  systems  of  so-called  ornamental  lighting. 

The  problem  of  laying  out  a  "great  white  way"  system  as 
described  above  is  quite  different  from  the  problem  of  lighting 
the  streets  of  the  residential  section,  for  the  latter  is  largely  a 
utilitarian  one  and  bears  on  the  matter  of  police  protection  and 
suitable  lighting  for  pedestrians,  motorists,  and  other  users  of  the 
thoroughfare. 

Residential  street  lighting  is  from  its  very  nature  of  low  average 
intensity.  As  this  class  of  lighting  comprises  a  relatively  large 
percentage  of  the  city's  streets,  it  is  important  for  obvious  rea- 
sons that  the  most  efficient  unit  suitable  be  used.  It  is  desirable, 
also,  to  employ  as  few  light  sources  as  possible  for  a  given 
average  intensity  of  illumination,  since,  the  apparent  affect 
produced  by  many  small  light  sources  is  that  of  a  very  much 
lower  intensity  of  illumination,  because  the  only  images  pro- 
duced on  the  retina  of  the  eye  are  those  of  the  light  sources 
themselves.  Especially  is  this  so  in  the  case  of  oiled  roadways 
where  the  amount  of  light  reflected  is  relatively  small. 

In  residential  street  lighting,  the  principle  of  silhouette  light- 
ing1 must  be  employed;  that  is,  seeing  is  accomplished  by  the 
discernment  of  objects  in  contrast  with  a  lighted  background, 
which  usually  is  the  street  surface,  rather  than  by  light  reflected 
from  the  objects  themselves. 

In  the  opinion  of  the  writer  a  non-uniform  illumination  is  more 
desirable  for  this  class  of  lighting  than  one  extremely  uniform, 
assuming  that  the  minimum  intensity  in  each  case  is  about  equal, 
but  the  average  intensity  higher  in  the  case  of  the  arc  lamps  than 
m  the  case  of  low  candle-powered  units ;  and  assuming,  of  course, 
the  same  expenditure  of  energy  per  linear  foot  of  street  and  that 
each  light  source  is  properly  screened  by  means  of  a  diffusing 
globe. 

The  proper  lighting  of  drives,  highways  and  parkways  requires 
the  use  of  a  specially  designed  arc  lighting  unit  which 
combines  white  light  with  maximum  efficiency  and  a  low 
maintenance  cost,  as  well  as  a  low  initial  cost  when 
compared    with    the    installation    of      many    small    units     of 

1  Preston  S.  Millar,  "An  Unrecognized  Aspect  of  Street  Illumination,  Trans.,  I.  E.  S., 
p.  546,  Vol.  V.  (Oct.  1910). 


HALVORSON  I      STREET   LIGHTING  91 

low  candle-power.  The  cost  of  installation  for  such  light- 
ing is  a  serious  factor,  for  there  is  usuall)  comparatively 
little  money  available  for  this  class  of  lighting.  The  prob- 
lem of  illumination  concerns  almost  wholly  the  motorist  and 
drivers  of  other  vehicles.  One  continually  reads  of  serious  night 
accidents,  involving  the  automobile  usually  with  a  horse-driven 
vehicle  or  a  motorcycle,  which,  statistics  show,  could  have  been 
avoided  had  adequate  illumination  been  provided.  The  require- 
ments of  such  illumination  are  not  greatly  different  from  those 
of  the  residential  streets,  as  described  above ;  that  is,  in  the 
employment  of  the  silhouette  principle  of  lighting.  With  such 
lighting  the  ability  of  the  eye  to  see  objects  clearly  a  sufficient 
distance  ahead  to  avoid  collision,  is  greater  than  with  any  other 
type  of  illumination.  Obviously,  it  is  necessary  that  the  light 
sources  should  be  mounted  well  above  and  outside  the  direct 
line  of  vision,  and  that  they  should  be  of  low  intrinsic  brilliancy, 
as  too  intense  a  light  source  destroys  the  adaptability  of  the  eye 
to  low  average  intensity  work. 

For  business  streets  and  other  thoroughfares  where  the  re- 
quirement is  the  very  best  kind  of  high-intensity  illumination,  the 
great  white  way  lamps*  may  be  used,  as  they  furnish  pearl- 
while  light  of  low  intrinsic  brilliancy,  produce  correct  light 
reflection  from  the  walls  of  varied-colored  buildings,  enhance 
the  window  display,  and  above  all,  attract  people  to  the  streets 
for  the  improvement  of  business. 

In  many  cases  a  satisfactory  spacing  for  units  of  this  type  has 

*  Since  the  commercial  introduction  of  the  ornamental  luminous  arc  lamp  at  New 
Haven,  in  December,  1911,  there  have  been  designed  three  new  forms  of  ornamental 
luminous  arc  lighting  units,  known  as  the  "Great  White  Way"  lamp  (two  forms)  con- 
suming 520  watts  at  6.6  amperes,  and  320  watts  at  4  amperes;  the  residential  lamp,  con- 
suming 300  watts  at  4  amperes;  and  the  parkway  lamp  designed  for  operation  at  both  4 
and  6.6  amperes.  300  and  500  watts  respectively.     These  lamps  are  shown  in  Fig.  1. 

The  great  white  way  lamps  are  for  use  on  the  principal  business  streets  and  cities. 
The  intensity  of  such  illumination  of  necessity  must  be  high  and  the  distribution  good. 
Therefore,  the  units  employed  should  be  placed  comparatively  close  together. 

The  residential  lamp,  as  its  name  implies,  in  addition  to  its  use  in  "Great  White 
Way"  lighting,  is  also  nsed  on  fine  residential  streets  and  on  boulevards  bordered  by 
large  estates,  where  as  a  rule,  shade  trees  overhang  the.  lamp  location.  In  such  cases 
the  low  mounting  of  the  light  source  (12  feet — 3.66  m.)  permits  good  illumination  as  it 
escape  screening  by  the  foliage. 

The  parkway  lamp  gives  a  somewhat  more  extended  light  distribution  than  the  two 
units  just  mentioned  and  is  designed  especially  for  roadways  where  an  extremely  low 
intensity  of  illumination  is  adequate  for  all  purposes.  It  is  usually  mounted  18  feet 
(5.486  m.)  above  the  roadway. 


92 


TRANSACTIONS  I.  E.  S. — PART  II 


been  found  to  be  approximately  75  to  100  feet  (22.86  to  30.48  m.) 
on  centers  on  each  side  of  the  street,  with  the  standards  in  a 
staggered  arrangement.    It  would  be  impossible  to  give  here  any 


pig.  1.— Ornamental  luminous  arc  lamps  and  standards  for  (A)  residental  streets, 
(B)  "great  white  way,"  and  (C)  parkway  lighting. 


definite  statement  on  the  exact  arrangement  to  be  followed  for 
future  installations,  as  obviously  each  case  must  be  considered 


halvorson:     street  lighting  93 

by  itself ;  the  character,  height  and  color  of  the  buildings,  as  well 
as  the  width  of  the  street,  all  playing  important  parts. 

For  residential  streets  and  broad  boulevards  where  the  screen- 
ing effect  of  foliage  must  be  considered  and  for  all  other  lighting 
classified  under  the  heading  "second  class,"  the  residential  lamp 
(see  A,  Fig.  i)  can  be  employed  to  give  the  best  results..  The 
spacing  of  this  lamp,  like  that  of  the  lamp  just  described,  varies 
with  the  local  conditions,  but  300  feet  (91.44  m.)  apart  in  stag- 
gered arrangement  gives  an  exceedingly  satisfactory  illumination. 

An  installation  of  this  character  will  show  fewer  light  sources 
within  the  range  of  vision,  beautify  the  streets  to  a  much  greater 
extent,  and  produce  a  more  satisfactory  illumination  for  the  same 
cost  of  installation  and  maintenance  than  any  other  lighting  unit 
available  at  the  present  time.  The  globes  used  with  these  lamps 
act  as  secondary  sources  of  pure  white  light  of  low  intrinsic  bril- 
liancy; consequently  the  units  themselves  are  extremely  pleasing 
to  the  eye  by  night  as  well  as  by  day,  fulfilling  all  esthetic  re- 
quirements. 

For  highways,  open  parks  and  drives  and  all  other  purposes 
where  an  efficient  ornamental  unit  is  desired  and  where  there  is 
but  little  danger  of  screening  the  light  source  by  foliage,  the 
parkway  lamp  (see  C,  Fig.  1)  may  be  employed  in  order  to  utilize 
to  the  highest  degree  the  principle  of  silhouette  lighting. 

With  these  three  units  it  is  believed  that  the  complex  and 
exacting  requirements  of  scientific  street  illumination  can  be 
successfully  met,  both  from  a  utilitarian  and  from  an  artistic 
viewpoint.  The  introduction  of  the  smaller  and  consequently 
lower  candle-powered  units  will  be  greatly  appreciated  by  those 
cities  which  possess  at  the  present  time  an  installation  of  the 
pendant  type  of  standard  luminous  arc  lamps,  as  these  new  orna- 
mental units  are  interchangeable  with  the  pendant  type  units 
insofar  as  their  operating  characteristics  are  concerned,  and  con- 
tain many  vital  mechanism  parts  such  as  magnets,  clutches,  etc., 
which  are  common  to  both.  This  feature  is  one  which  will  be 
greatly  appreciated  by  the  operating  man. 


94  TRANSACTIONS  I.  E.  S. PART  II 

DISCUSSION. 

P.  S.  Millar  (communicated)  :  Mr.  Halvorson's  paper  is 
somewhat  radical  in  that  it  departs  from  the  traditional  view  that 
small  illuminants  should  be  employed  on  side  streets  and  in 
residence  districts. 

Two  points  are  particularly  worthy  of  note,  although  some  may 
question  the  correctness  of  the  author's  view.  The  first  is  that 
comparatively  brilliant  illumination  of  a  thoroughfare  does  not 
detract  from  window  displays  if  there  is  considerable  contrast 
between  the  color  of  the  street  light  and  that  employed  in  the 
windows.  The  other  is  the  importance  of  the  appearance  which 
the  street  lighting  fixtures  present  in  the  daytime,  and  the  part 
which  they  play  in  the  general  appearance  of  a  street.  Both 
points  well  merit  greater  consideration  than  has  been  given  them 
in  the  past. 

The  silhouette  effect  in  street  lighting,  which  is  emphasized  so 
strongly  in  this  paper,  is  of  greatest  importance  where  the  re- 
quirements are  for  the  discernment  of  large  objects.  Lighting 
of  streets  with  few  large  units  rather  than  with  many  small  ones 
is  likely  to  promote  this  form  of  seeing.  It  must  not  be  for- 
gotten, however,  that  the  requirements  for  the  discernment  of 
small  irregularities  in  street  surface  are  important,  particularly 
to  the  pedestrian  and  the  driver  of  a  slow  moving  vehicle.  For 
any  given  street  and  lighting  appropriation  there  is  probably 
some  superior  balance  between  the  requirements  for  the  discern- 
ment of  large  objects  by  silhouetting  and  the  discernment  of 
street  surface  irregularities,  and  this  is  likely  to  imply  some 
balance  between  number  and  size  of  light  sources. 

The  illuminating  effect  is  one  of  a  number  of  considerations 
which  determine  the  choice  of  a  street  lighting  system.  Insofar 
as  it  is  the  leading  consideration,  our  understanding  of  street 
lighting  requirements  is  rather  vague.  Much  remains  to  be  done 
in  the  study  of  street  illumination  before  the  principles  which 
determine  this  balance  between  number  and  size  of  units  can  be 
announced.  Until  then,  practise  must  be  guided  by  conclusions 
drawn  from  views  of  individual  experts,  one  extreme  of  which 
appears  to  be  represented  in  this  paper.  It  is  the  logical  con- 
clusion of  this  extreme  view  that  in  all  classes  of  streets,  lamps 


STREET   LIGHTING  95 

which  are  of  relatively  large  power  may  be  employed  to  secure 
best  results.  There  will  probably  be  many  who  will  take  issue 
with  the  author  in  regard  to  this  conclusion.  It  is  to  be  hoped 
that  the  study  of  this  question  may  be  undertaken  seriously,  with 
a  view  of  answering  some  of  the  moot  questions  which  are  sug- 
gested by  this  paper. 

Mr.  W.  A.  Darrah  (communicated)  :  I  believe  that  the 
writer  has  given  a  very  clear  discussion  of  his  subject,  but  think 
that  perhaps  some  points  are  of  sufficient  importance  to  warrant 
a  further  discussion.  It  is  true  that  from  the  standpoint  of  the 
effect  produced  the  illumination  of  our  city  streets  is  a  matter 
which  cannot  be  fully  expressed  in  scientific  formulae  because  of 
the  psychological  element  involved.  There  are  certain  points, 
however,  which  are  capable  of  rather  exact  determination. 

Whether  the  lighting  of  residential  sections  is  considered  or 
whether  the  illumination  is  confined  entirely  to  that  of  business 
districts,  it  is  my  opinion  that  the  light  source  should  be  placed 
well  above  the  range  of  vision  in  all  ordinary  cases,  25  feet  to 
35  feet,  being  a  desirable  height.  In  fixing  upon  the  elevation  of 
the  light  sources,  consideration  should  be  given  to  the  illuminated 
signs  along  the  street  as  well  as  to  the  store  windows.  It  is 
obviously  undersirable  to  mask  the  effect  of  an  expensive 
illuminated  sign  by  an  arc  lamp  placed  closely  adjacent  to  it. 

I  believe  the  writer's  contention  that  a  light  source  of  relatively 
low  intrinsic  brillancy  is  desirable,  is  correct.  In  this  connection 
I  believe  the  flame  carbon  arc  is  a  decided  step  in  advance,  particu- 
larly when  used  with  diffusing  glassware.  Owing  to  the  relatively 
large  source  of  light  the  intrinsic  brilliancy  of  the  flame  arc  is 
considerably  less  than  that  of  the  older  types  of  arcs. 

A  further  consideration  of  the  subject  of  street  lighting  and 
one  which  has  not  been  discussed  in  the  present  paper  is  the  sub- 
ject of  station  apparatus.  In  deciding  upon  any  general  system 
of  lighting,  the  subject  should  be  considered  broadly  and  in  addi- 
tion to  the  specific  lamps  involved,  the  generating  and  accessory 
apparatus  should  be  considered.  The  simplicity  of  equipment 
and  low  cost  of  maintenance  of  alternating  current  apparatus  is 
a  further  fact  which  tends  to  emphasize  the  value  of  the  flame 
carbon  arc  lamp  as  a  unit  for  street  lighting. 


96  TRANSACTIONS  I.  E.  S. — PART  II 

The  flexibility  of  this  system  in  the  way  of  color  and  in- 
tensity is  an  additional  point  which  should  be  given  full  weight 
in  considering  a  street  lighting  system.  As  an  example  of  this 
flexibility  may  be  sighted  the  use  of  white  light  for  the  general 
illumination  of  business  streets  with  the  use  of  the  same  lamps 
equipped  with  carbons  for  producing  yellow  light  at  busy  street 
corners,  fire  alarm  boxes  and  other  points  which  require  special 
identification. 

J.  R.  Cravath  (communicated)  :  The  position  taken  by  the 
author  that  it  is  desirable  to  employ  as  few  light  sources  as 
possible  for  a  given  average  intensity  of  illumination  on 
residential  streets:  this  theory  is  obviously  subject  to  consider- 
able limitation  because  if  it  were  carried  to  its  logical  conclusion 
we  would  have  but  one  source  of  light  for  an  entire  length  of 
street. 

The  author's  opinion  that  non-uniform  illumination  is  more 
desirable  than  illumination  which  is  extremely  uniform  for  this 
class  of  lighting  is  open  to  considerable  question.  If  carried  to 
its  logical  conclusion,  this  would  drive  us  back  to  the  old  open  arc 
lamp  for  street  lighting,  and  we  would  have  to  discard  the  later 
types  of  lamps  such  as  the  magnetite  arc  which  gives  a  more  uni- 
form distribution  of  light  along  the  street.  Within  the  past  few 
years  considerable  intelligent  effort  has  been  directed  toward  pro- 
ducing more  uniform  illumination  along  a  street  on  the  theory 
that  the  extreme  contrast  between  the  areas  of  high  and  low 
illumination  is  detrimental  to  seeing  clearly,  as  is  also  the  blinding 
effect  of  high  candle-power  lamps  spaced  at  infrequent  intervals. 
There  is  no  doubt  whatever  that  changes  in  the  direction  of  better 
uniformity  have  resulted  in  a  decidedly  improved  street  lighting 
in  many  residence  streets.  But  there  is  some  question  how  far 
we  can  go  in  the  direction  of  greater  uniformity  to  advantage, 
and  the  point  raised  by  Mr.  Halvorson  is  a  very  interesting  one 
which  should  be  the  subject  of  thorough  investigation  before  con- 
clusions are  drawn  as  to  how  far  it  will  pay  to  go  in  securing 
uniformity  of  light  distribution. 

Prof.  J.  M.  Bryant  (communicated)  :  The  author  has  ap- 
parently made  a  careful  study  of  the  general  conditions  required 
for  lighting  in  a  medium  sized  city.     Too  many  cities  have  made 


STREET   LIGHTING  97 

the  mistake  of  putting  all  their  available  money  into  the  "great 
white  way"  because  they  have  obliged  to  appease  the  business 
men  who  are  influential  citizens  paying  a  considerable  part  of 
the  city  taxes. 

Within  the  last  few  years  another  portion  of  the  city,  namely 
the  better  residential  portions,  parks  and  boulevards  have  re- 
ceived more  attention  since  this  same  influential  citizen  has  found 
it  a  pleasure  to  be  out  in  the  evening  in  his  automobile.  As  yet 
the  poorer  business  and  manufacturing  portions  of  the  city  as 
well  as  the  poorer  residential  portions  have  received  but  little  con- 
sideration as  to  lighting.  Since  it  is  here  that  the  greatest  amount 
of  crime  exists  our  aim  should  be  to  light  and  care  for  these  dis- 
tricts in  a  much  better  manner,  in  order  to  improve  the  character 
of  the  inhabitants  and  to  aid  in  police  protection.  We  should  be 
as  ready  to  spend  the  city's  money  for  this  purpose  as  we  are  our 
own  for  charitable  or  police  work  among  these  inhabitants. 

No  one  form  of  lighting  unit  can  be  employed  with  equal 
effectiveness  or  economy  in  all  parts  of  a  city.  The  arc  lamp 
due  to  its  high  candle-power  must  be  used  for  lighting  in  places 
requiring  strong  light  and  also  rather  wide  distribution.  How- 
ever, due  to  the  relatively  high  coefficient  of  reflection  of  build- 
ings, this  type  of  unit  may  be  used  in  business  streets.  The  light 
reflection  from  buildings  tends  to  bring  up  the  low  intensity 
areas  and  make  the  illumination  uniform  if  the  lamps  are  not 
too  widely  spaced.  In  park  lighting  and  broad  boulevards  the 
arc  light  may  be  employed  to  advantage,  provided  enclosing 
globes  are  used,  thus  lowering  its  intrinsic  brilliancy. 

In  lighting  the  streets  in  residence  quarters  the  lower  candle- 
power  unit  obtained  only  in  incandescent  lamps  such  as  the  series 
tungsten  lamp  should  be  employed.  On  account  of  shade  trees 
and  other  features  these  lamps  must  be  hung  relatively  low, 
easily  coming  into  the  general  line  of  vision.  It  has  been  found 
advisable  to  screen  even  those  lamps  by  globes  to  make  it  easier 
for  the  eye  to  distinuish  distant  objects.  I  can  not  agree  with  the 
author  in  the  use  of  arc  lamps  widely  spaced  for  this  part  of  the 
installation. 


<\< 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 

Published  monthly,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIII 


MARCH.  1913 


No.  3 


Council  Notes. 

A  regular  meeting  of  the  council  was 
held  in  the  general  offices  of  the  society, 
jm  West  39th  Street,  New  York,  March 
14,  1913.  Those  present  were :  Preston 
S.  Millar,  president;  C.  O.  Bond,  Percy 
W.  Cobb.  Joseph  D.  Israel,  general  sec- 
retary; A.  E.  Kennelly,  V.  R.  Lansingh, 
Norman  Macbeth,  L,.  B.  Marks,  treas- 
urer; C.  J.  Russell  and  W.  J.  Serrill. 

An  oral  report  supplemented  by  a 
statement  of  the  society's  expenses  and 
membership  covering  the  first  two 
months  of  1913  was  received  from  the 
general  secretary. 

An  estimate  of  the  expenses  and  in- 
come of  the  society  for  the  fiscal  year, 
January  1  to  October  1,  1913,  was  re- 
ceived from  the  finance  committee.  The 
income  was  given  as  $8,870;  the  ex- 
penses $8,800. 

The  following  appropriations  were 
authorized:  $245  for  a  typewriter  and 
dictation  outfit;  $80  for  printing  and 
$20  for  a  membership  campaign  by  the 
Chicago  section. 

Payment  of  vouchers  Nos.  1210  to 
1240,  inclusive,  aggregating  $790.18  was 
approved  upon  recommendation  of  the 
finance  committee. 

Progress  reports  were  received  from 
the  committee  on  collegiate  education, 
the  committee  on  sustaining  member- 
ship, the  committee  on  section  develop- 


ment and  the  committee  in  charge  of 
the  organization  of  the  New  Lake  Erie 
section. 

In  the  report  of  the  section  develop- 
ment committee  the  desirability  of  es- 
tablishing a  student  grade  of  member- 
ship was  raised.  The  secretary  was 
directed  to  forward  to  the  committee 
on  collegiate  education  the  information 
and  data  on  the  subject  contained  in 
the  report. 

Monthly  reports  on  the  activities  of 
sections  were  received  from  the  follow- 
ing vice-presidents :  J.  R.  Cravath, 
Chicago  section;  J.  W.  Cowles,  New 
England  section;  W.  J.  Serrill,  Phila- 
delphia section;  Howard  S.  Evans, 
Pittsburgh  section. 

The  illumination  primer  committee 
reported  that  it  had  effected  a  few  minor 
changes  in  the  primer,  also  that  an 
order  for  100,000  copies  was  pending. 

A  progress  report  was  also  received 
from  the  committee  on  reciprocal  rela- 
tions with  other  societies. 

In  accordance  with  a  recommendation 
of  the  latter  committee  the  president 
was  directed  to  appoint  committees  to 
care  for  the  Illuminating  Engineering 
Society's  part  in  the  fourth  Interna- 
tional Congress  for  School  Hygiene  in 
Buffalo,  August  25  to  30,  1913,  and  the 
convention  of  the  American  Gas  Insti- 
tute at  Richmond,  Ya..  next  fall.  The 
arrangement  of  the  joint  meetings  which 


TRANSACTIONS    I.    E.    S. — PART    I 


will  be  held  on  the  occasion  of  both 
the  latter-mentioned  events  is  the  result 
of  activities  of  the  committee  on  recip- 
rocal relations  with  other  societies. 

It  was  decided  to  have  available  for 
joint  meetings  with  other  societies  1,000 
copies  of  the  primer  for  free  distribu- 
tion. It  was  understood  that  these 
copies  would  be  marked  "Compliments 
of  the  Illuminating  Engineering  Society" 
and  have  on  the  inside  a  notation  about 
the  work  of  the  society  and  the  prices 
of  the  primer. 

The  committee  appointed  to  devise 
plans  for  the  appointment  of  local  rep- 
resentatives in  cities  not  having  sections 
of  the  society  reported  that  it  had  con- 
sidered "the  question  of  proper  pro- 
cedure in  the  selection  of  local  repre- 
sentatives and  had  come  to  the  conclu- 
sion that  this  is  a  matter  in  which  each 
locality  presents  a  special  problem  by 
itself.  It  was  thought,  therefore,  that 
as  questions  of  local  representatives 
arise  they  should  be  handled  by  the 
general  officers  of  the  society  who 
should  consult  with  the  members  in  the 
locality  in  question  and  use  such  other 
means  as  may  seem  best  to  them  to 
arrive  at  a  proper  conclusion."  The 
report  was  accepted  and  the  committee 
discharged. 

It  was  resolved  that  the  advertising 
committee  should  be  instructed  to  com- 
plete negotiations  for  advertising  to  the 
extent  outlined  in  dieir  annual  report 
and  that  further  negotiations  for  ad- 
vertising in  excess  of  that  reported  be 
not  made. 

President  Millar  reported  upon  prog- 
ress of  plans  for  the  reorganization  of 
the  International  Photometric  Commis- 
sion. President  Vautier  of  that  Com- 
mission has  appointed  a  sub-committee 
consisting  of  representatives  of  the 
several  national  laboratories,  which  sub- 


committee has  been  asked  to  formulate 
plans  for  reorganization  of  the  Com- 
mission in  such  a  way  as  to  make  it 
thoroughly  representative  of  illumina- 
tion interests.  Such  plan  is  to  be  sub- 
mitted to  the  several  national  gas  socie- 
ties for  approval.  Dr.  E.  B.  Rosa  of  the 
Bureau  of  Standards  is  the  American 
representative  upon  the  sub-committee. 
Proposals  of  the  sub-committee  con- 
template for  this  country  the  issuance 
of  an  invitation  by  the  American  Gas 
Institute  to  other  national  societies  in- 
terested in  light,  illumination  and  pho- 
tometry to  meet  for  the  formation  of  a 
national  committee,  which  committee  is 
to  delegate  representatives  to  represent 
the  American  committee  at  the  first 
meeting  of  the  proposed  Commission, 
which  is  projected  for  Berlin  in  Sep- 
tember, 1913.  All  of  these  plans  are 
tentative  and  subject  to  change,  and  the 
ultimate  arrangement  may  be  materially 
different  from  that  suggested.  The 
present  status  of  the  work  of  the  sub- 
committee is,  however,  as  indicated. 

The  following  letter  was  received 
from  Mr.  Comfort  A.  Adams,  secretary 
of  the  Standards  Committee  of  the 
American  Institute  of  Electrical  Engi- 
neers : 

"At  the  last  meeting  of  the  Standards 
Committee,   it  was  voted  to  present  to 
the  Board  of  Directors  of  the  A.  I.  E.  E. 
the  following  by-law  for  their  adoption : 
'The    Standards    Committee   of    the 
A.  I.  E.  E.  is  instructed  by  the  Board 
of  Directors  of  the  A.  I.  E.  E.  to  take 
no  action  on  any  subject  matter  out- 
side   of    the    field    of    electrical    and 
magnetic    standardization    and    within 
the  field  of  the  Standards  Committee 
of    another    national    society,    before 
coming    to    an    agreement    with    the 
Standards  Committee  of  that  society, 
provided  that  the  said  society  instructs 


TRANSACTIONS    I.    E.    S. — PART    I 


its   Standards  Committee  not  to  take 
action  in  electric  or  magnetic  stand- 
ardization before  coming  to  an  agree- 
ment with   the   Standards   Committee 
of  the  A.  I.  E.  E.' 
"As  a  thirty  days'  notice  is  required 
before  the  vote  can  be  taken,  the  Board 
of    Directors    will    not    be    able    to    act 
upon  this  until  their  April  meeting.     In 
order    to    avoid    delay,    the    Standards 
Committee  urge  that  similar  action  be 
taken  by  your  society." 

It  was  resolved  that  the  committee  on 
nomenclature  and  standards  of  the 
Illuminating  Engineering  Society  be 
urged  to  adopt  a  resolution  similar  to 
that  contained  in  Mr.  Adam's  communi- 
cation. 

President  Millar  was  authorized  to 
appoint  a  committee  on  time  and  place 
for  the  1913  convention. 

President  Millar  announced  appoint- 
ments to  the  various  standing  and  tem- 
porary committees  of  the  society  which 
he  had  made  since  the  January  council 
meeting.  These  appointments  were 
approved. 


New  Members. 

The  following  applicants  were  elected 
members  of  the  society  at  a  meeting  of 
the  council  held  March  14.  1913: 
Argabrite.  H.  M. 

Manager,     Elwood     Electric    Light 
Company,  Elwood,  Ind. 
Bush.  W.  E. 

Illuminating  Engineer.  The  British 
Thomson-Houston  Company,  Ltd., 
J7  Upper  Thames  Street,  London. 
Eng. 

Hull,  Schuyler  M. 

Circuit  Breaker  Engineer,  Westing- 
house  Electric  &  Mfg.  Company, 
East  Pittsburgh,   Pa. 


Johnson,  C.  W. 

Salesman,  Westinghouse  Electric  & 
Mfg.  Company,  1205  Dime  Savings 
Bank  Bldg.,  Detroit,  Mich. 

Junes,  Geo.  A. 

Salesman.  Macbeth-Evans  Glass 
Company,  Pittsburgh,   Pa. 

Kilmer,  W.  S. 

Lighting  Engineer,  H.  W.  Johns- 
Manville  Company.  New  York.  N.  Y. 

Kirkpatrick,  R.  B. 

Salesman.  The  Philadelphia  Electric 
Company,  1000  Chestnut  Street, 
Philadelphia,  Pa. 

Kxauber,  Alex.   M. 

President,  Alex.  M.  Knauber  Com- 
pany, 742  Euclid  Avenue.  Oak  Park, 
111. 

Lyxch.  James  D. 

Chief  Engineer,  Lit  Brothers,  6th 
&  Market  Sts.,  Philadelphia,  Pa. 

Muxroe,  Roy  G. 

Service  Supervisor,  The  Denver  Gas 
&  Electric  Company.  Gas  &  Electric 
Bldg..  Denver,   Col. 

Pearsox,  Julius  T. 

Lamp  Salesman.  Westinghouse  Lamp 
Company,  27  Woodward  Avenue, 
Detroit.  Mich. 

Ross,  Jay   H. 

Secy.-Treas.,  American  Electrical 
Equipment  Co.,  Kansas  City,  Mo. 

Smullex.  R.  W. 

The  Philadelphia  Electric  Company, 
1000  Chestnut  Street.  Philadelphia. 
Pa. 

Young,  A.  W. 

Manager.  New  Business.  Public 
Service  Electric  Company.  418 
Federal   Street.   Camden,   N.    T. 


Sustaining  Members. 

The  following  companies  were  elected 
sustaining  members  of  the  society  at  a 
meeting  of  the  council  March  14,  1913: 


TRANSACTIONS    I.    E.    S. — PART    I 


Electrical  Testing  Laboratories. 
Holophane  Works  of  General  Electric 

Company. 
The     Edison     Electric     Illuminating 

Company  of  Boston. 
The  New  York  Edison  Company. 
The  Philadelphia  Electric  Company. 


Section  Activities. 

CHICAGO    SECTION 

At  a  regular  meeting  of  the  Chicago 
section  in  the  auditorium  of  the  Western 
Society  of  Engineers,  Monadnock  Block, 
March  12,  Mr.  Meyer  J.  Sturm  pre- 
sented a  paper  entitled  "Practical  Ideal- 
ism in  Illumination  with  Particular 
Reference  to  Hospitals."  The  paper 
will  appear  in  a  later  issue  of  the 
Transactions.  Eighty  members  and 
guests  were  present. 

NEW    ENGLAND    SECTION 

The  New  England  section  held  a  joint 
meeting  with  the  New  England  section 
of  the  National  Commercial  Gas  Asso- 
ciation in  the  auditorium  of  the  Edison 
Illuminating  Company  Building,  39 
Boylston  Street,  Boston,  March  12. 
Mr.  Preston  S.  Millar,  president  of  the 
I.  E.  S.,  presented  a  paper  on  "Some 
Phases  of  the  Illumination  of  Inte- 
riors" ;  the  paper  was  supplemented  by 
a  demonstration  of  lighting  effects  in 
miniature  rooms.  The  paper  and  dis- 
cussion appears  in  this  issue  of  the 
Transactions. 


NEW    YORK    SECTION 

The  New  York  section  held  a  meet- 
ing in  the  United  Engineering  Societies 
Building,  29  West  39th  Street,  New 
York,  March  13,  1913.  Three  other 
societies — New  York  Association  for 
the  Blind,  Committee  on  Prevention  of 


Blindness;  American  Society  of  Me- 
chanical Engineers,  American  Museum 
of  Safety — participated  in  the  meeting. 
Three  papers  were  presented :  "Illumina- 
tion and  Eyestrain"  by  Dr.  Ellice  M. 
Alger;  "Mechanical  Safety"  by  Dr.  W. 
H.  Tolman,  and  "Industrial  Lighting"  by 
Ward  Harrison.  The  first  one  of  these 
papers  appears  in  this  issue  of  the 
Transactions. 

philadelphia  section 
The  March  meeting  of  the  Philadel- 
phia section  was  held  on  the  28th  of 
the  month  at  the  Pennsylvania  Academy 
of  the  Fine  Arts.  Fifty  ladies  and 
gentlemen  were  present  at  the  dinner 
at  the  Hotel  Walton  preceding  the 
meeting.  Through  the  courtesy  of  the 
directors  of  the  Academy,  the  art  gal- 
lery was  open  for  inspection  before  the 
meeting,  and  this  proved  a  very  enjoy- 
able feature.  At  the  meeting  Mr.  M. 
Luckiesh  of  the  National  Electric  Lamp 
Association  presented  a  lecture  and 
demonstration  on  the  subject  of  "Light 
and  Art."  By  means  of  demonstrations, 
Mr.  Luckiesh  showed  the  effect  of  di- 
rection, color  and  quantitative  distribu- 
tion of  lights  on  objects  of  architecture, 
sculpture  and  paintings.  About  140 
members  and  guests  were  present. 

PITTSBURGH    SECTION 

A  meeting  of  the  Pittsburgh  section 
was  held  in  the  auditorium  of  the 
Engineers'  Society  of  Western  Pennsyl- 
vania, Oliver  Building,  March  14,  1913. 
Mr.  M.  Luckiesh  of  the  National  Elec- 
tric Lamp  Association,  Cleveland,  O., 
presented  a  paper  entitled  "Light  and 
Art"  which  was  supplemented  by  a 
series  of  demonstrations  showing  the 
effect  and  influence  of  quality  and  direc- 
tion of  light  on  various  art  objects. 
Thirty  members  were  present. 


TRANSACTIONS 


OF  THE 


Illuminating 
Engineering  Society 


MARCH,  1913 


PART  II 


c^ 


f% 


Papers,  Discussions  and  Reports 


[  MARCH,  1913  ] 

CONTENTS  -  PART  II 


Some  Phases  of  the  Illumination  of  Interiors.     By  Preston 
S.  Millar 99 

Illumination  and  Eyestrain.     By  Ellice  M.  Alger 130 

Home  Illumination.     By  George  Leland  Hunter 149 


11 


SOME  PHASES   OF   THE  ILLUMINATION   OF 

INTERIORS.* 


BY   PRESTON   S.    MILLAR. 


Synopsis:  This  paper  is  an  exposition  of  some  of  the  fundamentals 
of  interior  lighting.  It  treats  of  the  questions  of  glare,  diffusion,  direction 
of  light,  contrast,  influence  of  colored  surroundings  on  illumination,  etc., 
and  describes  and  illustrates  some  of  the  ordinary  methods  of  lighting 
now  in  use.  The  introduction  presents  a  brief  discussion  of  the  work 
and  functions  of  the  Illuminating  Engineering  Society  and  the  illuminating 
engineer. 

I.— THE  ILLUMINATING  ENGINEERING  SOCIETY  AND  ITS 
RELATION  TO  PUBLIC  LIGHTING  COMPANIES. 

The  business  of  the  lighting  company  is  to  sell  illuminating, 
power  and  heating  service.  Its  aims  and  achievements  are 
broader  than  the  mere  selling  of  electrical  or  gas  energy.  In 
the  conduct  of  its  business,  the  well  organized  large  lighting 
company  must  have  on  its  staff,  power  experts,  heat  and  ventila- 
tion experts,  and  illumination  experts.  The  illumination  expert 
should  be  well  qualified  to  deal  with  lighting  problems,  and  to 
the  extent  that  his  services  are  applied  on  behalf  of  the  com- 
pany and  its  customers,  the  lighting  company  may  be  said  to  be 
engaged  in  illuminating  engineering  work.  Such  activity,  how- 
ever, forms  but  an  insignificant  part  of  the  company  enterprise. 
While  its  importance  is  more  generally  recognized  than  for- 
merly, such  illuminating  engineering  of  necessity  is  restricted 
very  largely  to  the  application  of  existing  knowledge,  rather 
than  to  the  development  of  new  knowledge  in  the  field  of  illumi- 
nation. The  lighting  company  applies  developments,  and  it  is 
only  natural  that  the  art  of  illumination  (art  in  the  sense  of 
application  of  knowledge)  should  be  emphasized  in  lighting  com- 
pany illuminating  engineering,  and  that  the  development  of  the 

*  A  paper  and  demonstration  given  before  the  following  sections  of  the  Illuminating 
Engineering  Society  ;  New  Vork,  November  iS,  1912,  Philadelphia,  February  21,  1913, 
New  England,  March  12,  1913. 


IOO  TRANSACTIONS  I.  E.  S. — PART  II 

science  of  illumination  should  receive  but  little  impetus  therefrom. 

Thus  associations  of  lighting  companies  are  interested  in  the 
practise  of  illuminating  engineering.  And  the  discussion  of 
improvements  in  lighting  practice  is  a  feature  of  increasing  inter- 
est in  the  meetings  of  such  associations.  There  are,  however, 
other  aspects  of  illumination  not  properly  included  in  these  pro- 
ceedings which  may  not  be  neglected  if  the  science  and  art  are 
to  be  developed  properly. 

The  problem  of  illuminating  engineering  may  be  summed  up 
in  a  practical  way  as  follows :  Illumination  must  be  provided 
with  a  view  to  rendering  visible  the  things  which  it  is  desired 
to  have  seen.  The  illuminating  engineer,  in  studying  each  prob- 
lem, must  ascertain  what  it  is  desired  to  have  seen,  and  for  this 
purpose,  must  inform  himself  concerning  the  requirements  and 
the  viewpoints  of  those  who  are  furnishing  the  lighting,  and 
those  who  are  expected  to  see  by  its  aid.  In  a  machine  shop,  the 
illuminating  engineer  must  put  himself  in  the  place  of  the 
mechanic,  and  must  so  design  the  installation  that  the  mechanic 
can  see  the  surfaces  upon  which  he  works  and  can  see  to  apply 
tools  properly.  In  the  building  in  which  the  architect  has  sought 
for  certain  effects,  the  illuminating  engineer  must  provide  light 
with  which  to  display,  in  a  proper  way,  those  surfaces  and  orna- 
mentations which  the  architect  desires  to  have  seen.  Briefly,  the 
illumination  must  be  designed  for  the  particular  purpose  for 
which  it  is  to  be  used. 

The  illuminating  engineer  must  not  only  render  visible  the 
things  which  are  to  be  seen;  he  must  also  establish  and  maintain 
hygienic  conditions  for  the  eyes  and  body.  It  must  be  prac- 
ticable to  see  without  injury  to  the  eyes,  and  without  discomfort. 
It  may  be  that  in  the  proper  discharge  of  this  part  of  his  func- 
tions, the  illuminating  engineer  may  be  called  upon  to  go  beyond 
his  strict  province,  and  to  influence  conditions  other  than  those 
of  illumination.  For  example,  he  may  have  to  urge  the  use  of 
suitable  paper  in  schoolbooks. 

The  illuminating  engineer  must  not  only  render  visible  the 
things  which  are  to  be  seen,  and  make  vision  possible  under 
hygienic  conditions ;  he  must  also  consult  esthetic  requirements, 
and  conform  to  correct  principles  of  architecture  and  decora- 
tion, thereby  satisfying  discriminating  taste.     He  must   choose 


MILLAR:      ILLUMINATION  OF  INTERIORS  IOI 

fixtures  and  lighting  equipment  which  will  be  in  harmony  with 
the  character  of  the  installation  and  the  decorations.  He  must 
so  distribute,  diffuse,  and  modify  the  color  of  the  light,  as  to 
produce  pleasing  effects. 

These  three  requirements  having  been  met — a  tasteful  and 
satisfying  installation  having  been  provided,  with  which  it  is 
possible  to  see  with  comfort  the  things  which  it  is  desired  to  have 
seen — the  illumination  may  be  said  to  be  effective,  and  the  work 
of  the  illuminating  engineer  may  be  said  to  have  attained  one 
of  its  primary  objects.  There  remains  the  important  and  funda- 
mental consideration  that  these  things  must  be  accomplished 
with  reasonable  economy.  If  large  energy  consumption  must 
be  incurred  in  order  to  make  the  installation  effective,  the  illumi- 
nating engineer  does  not  hesitate,  for  the  installation  is  efficient 
nevertheless.  It  cannot  be  efficient  unless  it  achieves  the  purpose 
for  which  it  is  designed,  and  large  energy  consumption  or  high 
maintenance  cost  does  not  necessarily  imply  inefficiency.  But  it 
is  found  usually  that  there  is  ample  opportunity  for  the  expert- 
ness  of  the  illuminating  engineer  to  manifest  itself  in  so  design- 
ing the  installation  that  it  shall  be  effective  in  accomplishing  the 
purpose  for  which  it  is  intended,  while  reducing  the  cost  very 
materially  below  that  which  would  be  required  in  order  to  secure 
the  same  lighting  effects  by  inexpert  methods. 

Here,  then,  is  the  illumination  art  in  a  nutshell :  to  render 
visible  the  things  which  it  is  desired  to  have  seen;  to  establish 
hygienic  conditions  for  vision;  to  conform  to  esthetic  require- 
ments; and  to  accomplish  these  things  with  reasonable  economy. 

It  is  conceivable  that  in  designing  a  building  those  responsible 
might  desire  to  provide  an  illuminating  equipment  which  should 
be  as  nearly  as  possible  perfect.  Having  investigated  the  prob- 
lem, it  might  be  determined  to  retain  a  group  of  specialists  in 
the  various  professions  which  are  concerned  with  the  problem. 
It  is  conceivable,  then,  that  there  would  be  retained  an  engineer 
who  is  conversant  with  the  properties  of  illuminants ;  a  con- 
tractor who  would  install  the  equipment;  a  fixture  designer  com- 
petent to  design  or  select  suitable  lighting  fixtures ;  a  glass  expert 
competent  to  produce  such  quality  of  glassware  as  it  might  be 
desired  to  employ  in  the  installation;  a  physicist  qualified  to 
apply  optical  laws  in  the  design  of  reflecting  and  diffusing  sur- 


102  TRANSACTIONS  I.  E.   S. — PART  II 

faces ;  an  accomplished  decorator ;  a  psychologist  and  an  ophthal- 
mologist, who  would  pass  upon  the  conditions  from  the  viewpoint 
of  vision  conservation.  These  specialists,  in  co-operation  with 
the  architect  of  the  building,  would  meet  to  decide  upon  plans  for 
providing  illumination.  The  viewpoints  of  each  would  be  impor- 
tant to  an  extent  that  would  make  neglect  seriously  prejudicial 
to  the  success  of  the  lighting  installation.  Each  specialist  might 
be  expected  to  know,  or  to  think  that  he  knows,  what  kind  of 
illumination  is  necessary  in  order  to  fulfill  the  requirements  from 
his  viewpoint,  but  he  would  not  know  how  to  obtain  such  illumi- 
nation. Furthermore,  each  specialist  would  be  more  or  less 
ignorant  of  the  illumination  requirements,  judged  from  the 
standpoints  of  the  other  members  of  the  committee.  Whether  or 
not  such  a  committee  of  experts  would  be  able  to  agree  upon  a 
particular  plan  for  illuminating  the  building,  can  be  left  to  the 
imagination. 

The  well-qualified  illuminating  engineer  must  be  informed  in 
regard  to  the  underlying  principles  of  each  science  and  art  which 
would  be  represented  by  a  specialist  on  our  hypothetical  com- 
mittee. It  is  too  much  to  expect  that  he  should  be  an  expert  in 
each  of  these  lines  : 

"A  man  so  various  that  he  seems  to  be 
Not  one,  but  all  mankind's  epitome" 

He  must,  however,  have  sufficient  knowledge  of  the  fundamentals 
and  sympathy  with  the  aims  of  each  of  the  sciences  and  arts  in- 
volved, to  bring  his  work  into  harmony  with  them,  and  it  may  be 
noted  that  he  should  be  better  qualified  than  would  the  committee, 
to  produce  a  well-balanced  design,  because  his  knowledge  of  the 
other  phases  should  enable  him  to  give  each  its  proper  weight  in 
final  consideration  of  the  subject. 

To  state  the  function  of  the  illuminating  engineer,  is  to  indicate 
the  work  of  the  Illuminating  Engineering  Society.  It  seeks  to 
be  the  forum  where  specialists,  engaged  in  each  of  the  sciences 
and  arts  which  enter  into  illuminating  work,  can  meet  for  study 
and  discussion  of  the  problem,  exchanging  views,  learning  from 
one  another,  and  endeavoring  to  establish  correct  principles  upon 
which  illuminating  engineering  must  be  based.  Once  established, 
through  the  meetings  and  Transactions  of  the  Society  these 


millar:     illumination  of  interiors  103 

principles  are  quickly  disseminated  among  the  membership,  and 
are  applied  in  the  practical  work  of  illuminating  engineers.  When 
proven  beyond  peradventure,  effort  is  made  to  make  them  known 
to  the  public,  as  in  the  case  of  the  Illumination  Primer,  recently 
published  by  the  Illuminating  Engineering  Society. 

This  brief  discussion  should  serve  to  indicate  the  character  of 
the  Illuminating  Engineering  Society.  There  is  a  real  need  for 
an  organization  which  shall  make  a  specialty  of  illumination, 
developing  the  science  and  the  art  so  that  it  may  be  applied  by 
lighting  companies  and  others.  I  believe  that  an  impartial  study 
of  illumination  developments  of  the  past  six  years  will  lead  to  the 
conclusion  that  the  Illuminating  Engineering  Society  has  justified 
itself  by  its  achievements  and  by  its  present  status. 

II.— DEMONSTRATION  OF  LIGHTING  EFFECTS. 

It  is  a  peculiarity  of  illuminating  engineering  that  the  demands 
are  for  the  highest  technical  knowledge  and  skill  applied  in  the 
common  walks  of  life ;  that  success  or  failure  affects  closely  the 
people  in  their  ordinary  occupations.  The  most  technical  dis- 
cussion of  subjects  pertaining  to  illumination  is  likely  to  have  a 
practical  application,  of  interest  to  the  man  in  the  street.  It  is 
the  purpose  of  this  presentation1  to  indicate  in  a  non-technical 
manner  some  of  the  features  of  the  illumination  of  interiors 
which  have  been  studied,  and  upon  which  we  have  some  informa- 
tion as  a  basis  for  practise. 

If  a  room  be  illuminated  by  a  bare  lamp  (Fig.  1,  right),  the 
results  are  unsatisfactory  for  a  number  of  reasons.  In  the  first 
place,  the  walls  receive  the  major  amount  of  the  light  produced 
and  the  portions  of  the  room  in  which  the  light  is  more  likely  to 

1  For  the  lecture  upon  which  these  notes  are  based  miniature  rooms  were  constructed. 
These  were  4  by  4  feet  and  3^  feet  high.  Wall  decorations,  as  well  as  lighting  equip- 
ments, could  be  altered  readily.  With  one  or  two  exceptions  lamps  were  so  operated  as 
to  produce  64  lumens  in  each  room.  This  permitted  of  comparison  of  various  lighting 
systems  on  an  equitable  basis.  With  this  demonstration  equipment  a  wide  variety  of 
lighting  effects  could  be  produced  and  approximately  forty  were  projected.  Time  limita- 
tion, however,  restricted  those  actually  presented  to  twelve  which  are  here  illustrated  by 
photographs  available  through  the  courtesy  of  The  New  York  Edison  Company. 

The  temptations  to  deal  with  vaiions  types  of  interiors  and  to  undertake  studies  of 
color  were  resisted.  Only  one  type  of  interior,  and  that  in  its  simplest  form,  was  consid- 
ered, time  lacking  for  a  more  extended  discussion. 

The  suite  of  miniature  rooms  was  constructed  under  the  supervision  of  Mr.  W.  F. 
Little  and  was  operated  under  his  direction  by  Messrs.  H.  Bardwell,  M.  D.  Beuick  and  W 
Ihlefeld,  all  of  the  Electrical  Testing  Laboratories. 


104  TRANSACTIONS  I.  K.  S. — PART   II 

be  utilized,  are  inadequately  illuminated.  The  light  source  is  un- 
attractive, and,  when  within  the  field  of  vision,  is  annoying,  if 
not  actually  injurious  to  the  eyesight.  This  latter  effect,  included 
under  the  name  of  glare,  is  very  noticeable  in  the  illustration, 
where  the  lamp  is  at  the  center  of  the  field  of  vision,  and  the  effect 
is  exaggerated  beyond  that  which  would  be  experienced  by  oc- 
cupants of  the  room. 

If  the  lamp  be  shielded  from  view  (Fig.  I,  left),  conditions  are 
much  improved.  Much  of  the  discomfort  and  annoyance  dis- 
appears. While  the  distribution  of  light  on  surfaces  seen  within 
the  room  is  not  changed  materially,  yet  everything  can  be  seen 
more  distinctly.  Observe,  for  example,  the  vertical  stripes  upon 
the  wall-paper.  Beginning  near  the  floor,  trace  a  vertical  stripe 
which  is  at  some  height,  almost,  if  not  quite  in  line  with  the  lamp 
in  room  No.  i  (to  the  right).  It  will  be  noticed  that  as  the  gaze 
approaches  the  vicinity  of  the  lamp,  it  is  difficult  to  see  the  stripe, 
and  that  when  level  with  the  lamp,  the  stripe  disappears  entirely. 
In  room  No.  3  (to  the  left)  in  which  the  lamp  is  shielded  from 
the  eye,  the  corresponding  stripe  may  be  distinguished  when  look- 
ing just  past  the  lamp  and  screen. 

One  of  the  important  functions  of  a  reflector  or  other  lighting 
auxiliary  is  to  thus  shield  the  lamp  from  view,  by  interposing  be- 
tween it  and  the  eye,  either  an  opaque  or  a  translucent  medium. 
This  is  accomplished  in  room  No.  2  (Fig.  1,  middle). 

But  a  reflector  should  fulfill  other,  equally  useful,  purposes. 
In  shielding  the  lamp  from  view,  it  may  also  be  made  to  direct  a 
considerable  proportion  of  the  light  where  it  can  be  utilized  to 
best  advantage.  Much  study  has  been  given  to  this  aspect  of  the 
problem,  and  the  performance  of  any  standard  type  of  reflector 
may  be  ascertained  by  reference  to  photometric  tests  of  light  dis- 
tributed downward  may  be  judged  from  the  floor  brightness, 
day  prepared  to  supply  with  their  wares.  Perhaps  in  no  branch 
of  illumination  have  such  great  strides  been  made  in  the  past  ten 
years  as  in  the  design  of  reflectors  in  particular,  and  lighting 
auxiliaries  in  general. 

It  has  been  noted  that  the  bare  lamp  distributes  but  a  small  pro- 
portion of  the  light  downward.  In  room  No.  2  (Fig.  2)  a 
reflector  is  employed  which  re-directs  downward  a  goodly  propor- 


MILLAR:      ILLUMINATION  OF  INTERIORS  105 

tion  of  the  light,  illuminating  the  card  below  it  much  more 
brightly  than  would  a  bare  lamp.  In  room  Xo.  3,  this  re-direction 
of  light  is  effected  in  such  a  way  as  to  concentrate  a  large  propor- 
tion directly  below  the  lamp,  thereby  illuminating  the  card  to  a 
brightness  which  is  about  twice  that  of  the  card  in  room  No.  2, 
which  was  considerably  brighter  than  the  card  in  the  room  where 
no  reflector  was  used. 

In  room  No.  1,  as  now  equipped,  a  reflector  is  employed  which 
has  been  designed  without  regard  to  optical  laws,  and  which, 
though  looking  like  a  prismatic  reflector,  has  in  fact  almost  none 
of  the  qualities  which  characterize  such  glassware.  It  accom- 
plishes little  in  the  way  of  re-direction  of  light,  while  affording 
but  an  ineffectual  protection  for  the  eyes  against  brightness  of  the 
filament.  It  absorbs  a  certain  amount  of  light  without  rendering 
any  adequate  return  in  improvement  of  conditions. 

In  the  three  rooms  there  is  illustrated  the  range  of  practicable 
accomplishment  in  the  employment  of  reflectors,  if  we  omit 
opaque  reflectors,  which  would  not  be  suitable  for  employment 
under  such  conditions.  In  room  No.  1,  general  distribution  of 
the  light  throughout  the  room ;  in  room  No.  2,  effective  re-direc- 
tion of  much  of  the  light  downward,  largely  increasing  the 
intensity  on  the  table  plane,  though  illuminating  the  walls  and 
ceiling  brightly  enough  to  avoid  the  appearance  of  dimness.  In 
room  No.  3,  the  concentration  within  a  small  area  beneath  the 
lamp  is  very  marked,  this  being  effected  by  taking  from  the 
walls  and  from  the  table  plane  near  the  walls,  a  portion  of  the 
light  which  falls  upon  them  in  room  No.  2,  and  concentrating 
it  upon,  or  near  the  table.  The  relative  intensities  of  light  dis- 
tributed downward  may  be  judged  from  the  floor  brightness. 

The  correct  design  of  a  reflector  to  accomplish  a  given  purpose, 
involves  the  application  of  well-known  optical  laws.  With  pris- 
matic glass  and  mirror  types  of  reflectors,  a  wide  variety  of 
distribution  may  be  obtained.  With  opal  or  phosphate  glasses, 
such  as  that  in  room  No.  2,  the  possibilities  of  securing  high 
concentration  are  rather  more  limited,  though  with  this  one 
exception  these  too  may  be  designed  to  produce  practically  any 
distribution  likely  to  be  required. 

In  achieving  the  particular  distribution  which  characterizes  a 
given  reflector,  it  is  important  that  the  light  source  be  correctly 


106  TRANSACTIONS  I.  E.   S. — PART  II 

located  with  reference  to  the  reflector.  The  use  of  an  incorrect 
shade  holder,  or  of  an  improper  lamp  distorts  the  distribution 
and  usually  detracts  from  the  appearance  and  usefulness  of  the 
lighting  unit. 

In  reflectors,  as  well  as  in  globes  and  other  forms  of  glass 
lighting  auxiliaries,  the  degree  of  optical  density  is  important, 
affecting  both  the  performance  and  the  appearance  of  the  glass. 
This  is  an  important  feature  to  be  considered  in  selecting  glass- 
ware. In  the  now  rather  common  forms  of  display  street  light- 
ing, which  utilize  clusters  of  tungsten  lamps  in  globes,  very  dis- 
pleasing effects  are  sometimes  encountered,  due,  first,  to  the  non- 
uniformity  of  the  globes,  and  second,  to  the  insufficient  density 
which  makes  the  location  of  the  lamp  apparent,  instead  of 
rendering  the  whole  surface  of  the  globe  equally  bright,  making 
it  appear  a  ball  of  light.  Much  of  the  lighting  glassware  in  use 
in  residences  a  few  years  ago,  and  it  is  to  be  feared  even  to-day, 
consists  of  etched  or  frosted  crystal  glass  which  serves  chiefly 
to  give  the  fixture  a  somewhat  finished  appearance.  It  neither 
directs  sufficient  light  usefully  to  make  it  efficient,  nor  conceals 
the  light  source  sufficiently  to  make  it  attractive  or  of  value  in 
protecting  the  eyes. 

In  Fig.  3  there  is  a  globe  of  crystal  glass,  roughed  inside,  a 
light  opal  globe,  and  a  denser  opal  globe.  The  last  presents  a 
better  appearance  without  involving  serious  sacrifices  otherwise. 
The  light  absorptions  of  these  balls  are  respectively : 

Frosted  ball 6% 

Light  opal  ball 13% 

Dense  opal  ball 22  % 

\Yhen  employed  in  the  miniature  rooms  shown  in  Fig.  2  the 
relative  light  intensities  throughout  the  table  plane  averaged: 

Frosted  ball 100% 

Light  opal  ball 106% 

Dense  opal  ball 95  % 

In  passing  I  wish  to  refer  to  the  great  diversity  of  lighting 
fixtures  and  glassware  which  are  now  available  in  standard 
types,  awaiting  selection.  Practically  all  ordinary  requirements 
in  interior  illumination  may  be  filled  by  lighting  auxiliaries 
selected  from  those  now  upon  the  market.  In  regard  to  effi- 
ciency, most  of  the  reflectors  and  globes  which  pretend  to  be 


f 


Room  3. 


Room  2.  Room  1. 

Fig.  1. — Showing  importance  of  shading  lamp. 


Room  3.  Room  2.  Room  1. 

Fig.  2.  — Showing  variety  of  light  distributions  which  may  be  obtained  by  use  of  ordinary  reflectors. 


Light.  Medium.  Dense. 

Fig-  3- — Showing  improved  appearance  when  glassware  is  dense  enough  to  conceal  lamp. 


Room  3.  Room  2.  Room  1. 

Fig.  4.— Showing  appearance  with  various  wall  decorations. 


Room  3.  Room  2.  Room  1. 

Fig.  5. — Showing  appearance  with  various  wall  decorations. 


Room  3. 


Room  2. 
Fig.  10. — Indirect  lighting 


MILLAR:       ILLUMINATION  OF  INTERIORS  IO7 

efficient  accomplish  their  purpose  admirably.  In  fact  so  care- 
fully has  this  element  of  the  question  been  studied,  that  an  ineffi- 
cient reflector  cannot  to-day  be  successful  unless  it  has  some 
compensating  advantage,  which  renders  it  superior  for  some 
purposes  in  spite  of  its  inefficiency.  I  recently  went  through  the 
interesting  experience  of  making  a  comparison  of  the  standard 
types  of  reflectors  which  were  upon  the  market  ten  years  ago, 
and  comparing  them  with  types  now  available.  In  regard  to 
efficiency,  the  improvement  has  been  very  marked.  Absorptions 
of  10  to  20  per  cent,  now  rule,  where  ten  years  ago  absorptions 
of  25  to  40  per  cent,  were  typical  for  reflectors  of  substantially 
similar  light  distribution  characteristics.  Improvement  in  the 
quality  of  reflecting  surfaces  has  gone  hand  in  hand  with  im- 
provement in  the  design  of  the  curvature  of  such  surfaces.  As 
in  efficiency,  so  it  is  in  appearance.  Most  of  the  reflectors  of  ten 
years  ago  were  opaque,  and  few  were  pleasing  to  the  eye.  To-day 
even  in  the  reflectors  where  efficiency  in  light  re-direction  is  the 
chief  aim,  pleasing  appearance  is  the  accepted  order. 

In  the  class  of  lighting  auxiliaries  in  which  decorative  effect 
is  the  chief  object,  a  wide  variety  is  available,  and  much  of  it 
it  pleasing  and  tasteful.  Unfortunately,  however,  such  auxili- 
aries are  characterized  by  inefficiency  to  an  extent  which  appears 
rather  unnecessary.  It  is  probable  that  in  the  developments  of 
the  next  few  years,  we  shall  note  a  strong  tendency  to  improve 
the  efficiency  of  some  types  of  decorative  reflectors,  without 
interfering  with  their  decorative  qualities. 

The  influence  of  room  decoration  upon  the  amount  of  light  re- 
quired to  illuminate  a  room  properly  is  very  marked ;  or,  stated 
otherwise,  with  a  given  amount  of  light  produced  in  a  room,  the 
effectiveness  of  the  illumination  is  largely  influenced  by  the  char- 
acter of  the  decorations.  Considering  the  simple  case  of  a  bare 
lamp,  employed  to  illuminate  rooms  having  light,  medium  and 
dark  walls  respectively,  we  may  note  a  number  of  interesting 
effects  (Fig.  4).  In  the  first  place,  the  illuminated  card  on  the 
table  in  room  No.  3  appears  brighter  than  the  cards  in  the  other 
rooms.  It  must  be  apparent  that  the  card  cannot  be  brighter  be- 
cause it  receives  light  from  the  lamp  and  the  ceiling  only,  while 
the  card  in  room  No.  1,  for  example,  receives  light  from  the 
corresponding  light  sources  and  ceiling  which  is  enhanced  con- 
2 


IOS  TRANSACTIONS  I.  E.  S. — PART   II 

siderably  by  light  reflected  from  the  walls.  The  card  in  room 
No.  i  is  actually  30  per  cent,  brighter  than  the  card  in  room  No. 
3.  That  it  does  not  so  appear  is  an  example  of  the  effect  of  con- 
trast, which  in  illumination  is  a  very  important  fundamental.  A 
corresponding  comparison  may  be  made  by  observing  the  upper 
part  of  the  wall  in  each  room,  where  again  the  white  paper  ap- 
pears brighter  in  room  No.  3  than  in  the  other  rooms.  Though 
actually  not  so  bright  as  the  white  surfaces  in  rooms  Nos.  1  and 
2,  these  surfaces  appear  brighter  in  room  No.  3  in  comparison 
with  the  dark  wall-paper  to  which  the  eye  naturally  adapts  itself 
more  or  less. 

In  room  No.  1,  portions  of  the  furniture  which  are  but  slightly 
illuminated,  as  legs  of  the  table,  stand  out  distinctly,  being  sil- 
houetted against  the  light  rear  wall.  In  room  No.  3,  so  small  is 
the  contrast  between  the  rear  wall  and  the  dimly  lighted  portions 
of  the  furniture,  that  it  is  difficult  to  discern  the  latter. 

The  glare  due  to  the  exposed  light  source  is  more  serious  in 
room  No.  3,  due  to  the  larger  contrast  between  the  light  source 
and  the  walls.  Shadows  of  the  furniture  against  the  walls  are 
very  prominent  by  contrast  in  room  No.  1,  in  spite  of  the  fact 
that  the  shaded  areas  are  more  brightly  illuminated  by  light  which 
is  generally  diffused  within  the  room. 

In  considering  this  photograph,  it  should  be  remembered  that 
no  light-directing  auxiliary  has  been  employed,  and  that  therefore 
a  larger  proportion  of  the  light  falls  upon  the  walls  than  good 
practise  would  dictate,  if  we  except  the  darker  room.  In  most 
installations  it  is  desirable  primarily  to  secure  the  proper  illumina- 
tion of  the  lower  part  of  the  room,  where  the  light  is  utilized,  the 
other  requirements  being  that  the  ceiling  and  walls  shall  be 
illuminated  sufficiently  to  make  the  effect  pleasing.  When  re- 
flectors are  used,  the  lighting  effect  of  ceiling  and  wall  decorations 
is  reduced  greatly,  if  the  reflectors  are  concentrating  in  character, 
and  reduced  slightly,  if  they  distribute  the  light  rather  broadly, 
about  the  lower  part  of  the  room. 

The  brightness  of  walls  is  an  important  element,  affecting 
ocular  comfort  probably  more  seriously  than  the  illumination  of 
the  table  plane.  Generalizing,  it  is  probably  the  best  rule  to  avoid 
extremes  of  wall  decoration,  whether  they  be  light  or  dark.     If 


MILLAR:      ILLUMINATION  OF  INTERIORS  IO9 

the  walls  are  of  high  reflecting  power,  it  is  important  to  so  direct 
most  of  the  lighting  that  the  amount  permitted  to  fall  upon  the 
walls  will  not  render  them  so  excessively  bright  as  to  be  trying 
to  the  eyes.  The  illuminating  engineer  cannot  control  wall  de- 
corations, but  he  can  control  the  light  produced  within  the  room, 
and  can  so  direct  it  as  to  secure  the  best  effects. 

In  the  next  photograph  a  reflector  which  directs  the  light  down- 
ward rather  largely  is  shown. 

This  detracts  from  the  brightness  of  the  upper  portions  of  the 
walls,  the  change  of  course  being  most  apparent  in  room  No.  I, 
where,  due  to  the  relatively  high  reflecting  power  of  the  wall- 
paper the  wall  was  brightest  in  the  last  photograph.  The  lower 
portions  of  the  walls  are  somewhat  brighter  than  when  bare 
lamps  were  employed.  Due  to  the  better  lighting  of  the  floor, 
the  lower  portions  of  the  table  and  chairs,  which  with  the  bare 
lamps  could  hardly  be  seen,  are  now  slightly  illuminated.  With 
this  installation  the  effect  of  the  ceiling  and  walls  is  lessened,  be- 
cause a  smaller  amount  of  light  is  permitted  to  fall  upon  them 
reducing  their  illuminating  power.  That  is  to  say,  when  a  suit- 
able reflector  is  employed,  the  table  plane  illumination  intensity  is 
more  nearly  independent  of  reflection  from  ceiling  and  walls,  and 
instead  of  relying  upon  the  latter  for  assistance  in  producing  use- 
ful illumination,  the  problem  is  simplified  to  one  of  rendering  the 
walls  bright  enough  to  produce  a  cheerful  appearance. 

It  has  been  shown  in  the  above  that  the  ceiling  and  wall  decora- 
tions, when  light  in  tone,  may  be  of  material  assistance  in  increas- 
ing the  illumination  intensity  on  the  table  plane.  It  may  be  argued 
as  the  corollary  of  this,  that  when  the  walls  are  dark  and  incap- 
able of  augmenting  the  table  plane  illumination  materially,  the  use 
of  reflectors  for  that  purpose  is  all  the  more  important. 

I  have  discussed  the  effect  of  glare  due  to  the  presence  of 
a  bright  light  source  within  the  field  of  vision.  This  effect  would 
be  almost,  if  not  quite  as  disturbing,  if  instead  of  having  a  lamp 
within  view,  an  image  of  the  lamp  were  to  be  seen  in  a  mirror. 
In  that  event,  the  effect  would  be  due,  not  to  the  presence  of  the 
light  source,  but  to  specular  reflection  of  the  light  from  the 
mirrored  surface.  It  is  perhaps  unfortunate  that  most  artificial 
surfaces  which  we  are  likely  to  view  are  sufficiently  glossy  or 


HO  TRANSACTIONS  I.  E.  S. — PART  II 

polished  to  partake  in  some  measure  of  the  qualities  of  a  mirror ; 
that  is,  to  reflect  light  specularly.  Some  surfaces  which  are  very- 
mat  and  free  from  gloss,  diffuse  the  light  so  generally  that  the 
specular  element  of  the  reflection  is  immaterial  for  most  pur- 
poses. But  in  most  paper  employed  in  books  and  magazines, 
there  is  a  considerable  element  of  specular  reflection,  and  this 
characteristic  is  responsible  for  much  of  the  difficulty  which 
demands  adroit  handling  by  the  illuminating  engineer  in  utilitarian 
lighting. 

Referring  to  the  demonstration  cards  (Fig.  6),  it  will  be 
noted  that  the  paper  and  letter  on  the  right  half  of  each  have 
glossy  surfaces,  while  those  on  the  left  have  diffusing  surfaces 
being  almost  totally  free  from  specular  reflection.  On  the  right 
half,  one  may  see,  when  in  line  with  the  direction  of  the  reflec- 
tion, a  distorted  image,  or  a  number  of  distorted  images,  of  the 
light  source,  much  as  though  he  were  viewing  the  source  through 
a  very  imperfect  mirror.  On  the  left  it  is  noted  only  that  the 
surface  is  illuminated  and  no  trace  of  an  image  of  the  light 
source  may  be  seen.  From  all  positions  the  letter  on  the  left 
half  of  the  card  may  be  seen.  From  a  particular  direction 
(right  photograph),  that  upon  the  right  half  of  the  card  can  be 
seen  only  with  great  difficulty,  if  at  all,  because  it  is  viewed  from 
the  direction  in  which  the  glare  is  manifested. 

No  small  part  of  the  dissatisfaction  with  illumination  installa- 
tions is  due  to  this  effect  of  glare  from  observed  surfaces.  The 
statement  may  be  ventured  also  that  no  small  part  of  trouble 
with  eyes  is  traceable  to  the  same  source.  There  are  three 
remedies :  one  is,  to  eliminate  glossy  surfaces  wherever  possible ; 
particularly  is  this  important  in  schoolbooks,  and  it  is  very  grati- 
fying to  know  that  serious  efforts  are  being  put  forth  with  a 
view  to  regulating  this  matter.  The  second  remedy  is  to  reduce 
the  brightness  of  light  sources  as  much  as  practicable  by  passing 
the  light  through  a  diffusing  medium  of  large  area,  or  by  reflect- 
ing it  from  a  diffusing  surface  of  large  area,  in  order  that  when 
specular  reflection  from  an  observed  surface  is  encountered,  the 
brightness  of  the  light  reflected  may  be  so  low  as  to  minimize 
the  difficulty.  The  third  remedy  is  to  so  locate  light  sources,  or 
to  so  locate  the  illuminated  surfaces  and  adjust  the  position  in 
working  or  reading,  that  the  direction  in  which  light  is  reflected 


MILLAR:      ILLUMINATION  OF  INTERIORS  III 

specularly  shall  not  be  toward  the  eyes.  All  three  of  these 
possible  remedies  should  be  kept  in  mind  and  applied  wherever 
practicable,  and  any  one,  or  a  combination  of  a  part  of  each  of 
the  three  can  be  made  effectual  in  reducing  the  trouble  to  a  point 
where  it  is  not  serious.  The  growing  appreciation  of  the  impor- 
tance of  this  element  of  illuminating  engineering  work  has  been 
the  distinguishing  feature  of  the  past  two  years  in  the  illumina- 
tion field. 

The  oil  lamp  has  in  recent  past  years  been  the  standard  of 
comparison  for  artificial  illumination.  Even  to-day  it  is  tradi- 
tional among  oculists  that  there  is  no  artificial  illuminant  which 
yields  a  light  so  free  from  detriment  to  the  eyesight  as  does  the 
oil  lamp.  It  is  therefore  of  interest  to  note  some  of  the  condi- 
tions under  which  the  oil  lamp  has  been  used.  Being  essentially 
a  small  illuminant,  and  both  self-contained  and  portable,  it  was 
natural  that  it  should  be  placed  close  to  the  object  viewed.  This 
entailed  locating  it  more  or  less  on  a  level  with  the  eyes,  and  so 
near  to  the  observer  that  shielding  the  former  became  a  matter 
of  natural  course.  In  that  fact  is  to  be  found  the  reason  for 
the  development  of  the  oil  lamp  shade.  Given  a  small  illuminant. 
shaded  for  the  protection  of  the  eyes,  there  was  no  condition 
under  which  visual  difficulties  could  be  experienced  unless  it 
were  attempted  to  read  with  the  book  in  the  illuminated  zone 
near  the  lamp  base,  with  the  reader  facing  the  lamp  (Room  i). 
Under  such  conditions,  glare  due  to  specular  reflection  from 
the  paper  might  be  detrimental  to  vision,  in  which  case  it  would 
be  so  immediately  apparent  that  instinctively  the  reader  would 
shift  his  position  or  the  lamp  slightly,  in  order  to  avoid  it.  With 
light  from  this  single  light  source  incident  upon  the  page  from 
a  direction  which  would  not  result  in  serious  glare  from  the 
paper,  and  with  the  flame  shielded  from  the  observer's  eyes,  the 
conditions  for  reading  or  other  work  were  comparatively  good. 
At  the  same  time,  the  old  oil  lamp  was  well  adapted  to  the  illumina- 
tion of  a  book  by  light  from  over  the  reader's  shoulder  (Room  3), 
one  of  the  best  positions  for  reading.  It  was  in  the  comparative 
freedom  from  misuse  of  the  oil  lamp,  and  the  conditions  which 
its  employment  made  natural,  that  the  relative  freedom  from 
harmful  effect  was  probably  found,  if  such  freedom  did  exist 
under  those  conditions,  which  is  a  point  that  has  not  been  estab- 


112  TRANSACTIONS  I.  E.  S. — PART  II 

lished.  Of  course,  no  matter  how  favorable  conditions  for 
vision  may  be,  it  is  difficult  to  prevent  a  certain  amount  of  care- 
lessness or  perversity  in  the  use  of  the  light  (Room  2).  Whether 
it  is  carelessness  which  induces  Mrs.  Lux  to  read  with  her  page 
dimly  illuminated  by  light  from  the  ceiling  and  wall,  while  facing 
the  light,  is  open  to  question.  Perhaps  she  may  feel  that  her 
appearance  is  more  attractive  with  the  light  full  upon  her  face. 

Wall  brackets  may  be  employed  with  good  effect  if  equipped 
discreetly.  For  utilitarian  purposes  they  are  of  value  chiefly  in 
providing  local  illumination.  It  is  very  difficult  to  light  a  room 
solely  from  wall  brackets  (Fig.  8).  The  light  cannot  be  dis- 
tributed satisfactorily  in  the  room  without  placing  light  sources 
immediately  within  the  range  of  vision.  Wall  brackets  find  best 
application  when  employed  in  rooms  in  which  the  main  illumina- 
tion is  provided  otherwise  and  the  brackets  are  equipped  with 
decorative  shades,  the  installation  serving  purposes  of  ornament 
rather  than  utility. 

Daylight,  being  that  under  which  the  human  eye  has  been 
evolved,  may  be  expected  to  possess  the  qualities  for  which  the 
eye  is  best  adapted.  Neglecting  other  differences  between  the 
natural  conditions  for  which  the  eye  is  adapted  and  the  artificial 
conditions  with  which  we  have  surrounded  it,  (such  as  the  change 
from  distant  to  near  vision  and  the  change  from  use  of  the  eye 
during  daylight  hours  only,  to  use  of  the  eye  for  almost  as  long 
a  period  during  the  hours  of  the  night,)  there  still  remain  cer- 
tain differences  between  artificial  light  and  daylight,  the  study  of 
which  forms  a  most  interesting  field  for  the  illuminating  engineer. 
Daylight  out-of-doors  is  the  standard  against  which  we  must 
compare  both  artificial  light  and  daylight  indoors,  for  the  day- 
light which  is  available  in  our  interiors  differs  materially  from 
that  out-of-doors  in  respect  to  quality,  intensity  and  direction. 
The  intensities  may  be  from  0.01  to  0.001  of  those  which  prevail 
out-of-doors  in  bright  sunlight.  The  quality  may  differ  not  only 
in  respects  which  are  not  perceptible  to  the  eye,  but  often  differs 
in  color  due  to  the  influence  of  the  absorption  of  colored  walls, 
etc.,  which  materially  alter  the  color  of  the  natural  light.  The 
direction  is  usually  quite  different.  In  regard  to  the  desirability 
of  such  direction  of  light  as  that  which  is  prevalent  in  interiors 


millar:     illumination  of  interiors  113 

illuminated  by  daylight,  there  is  considerable  discussion  at  the 
present  time,  pro  and  con.  My  own  view  is  that  usually  the 
direction  is  undesirable.  Coming  through  a  window  or  windows 
on  one  side,  or  at  the  most,  two  sides  of  a  room,  usually  at  an 
angle  somewhere  between  the  horizontal  and  45  deg.  above  the 
horizontal,  the  light  is  very  unequally  distributed.  The  floor  and 
opposite  wall  receive  the  greater  part  of  it,  while  the  wall  on  the 
side  of  the  room  in  which  the  windows  are  cut,  is  illuminated  only 
by  such  light  as  may  be  reflected  from  the  floor  and  the  opposite 
wall.  Persons  sitting  in  the  room  are  likely  to  have  the  window 
and  the  bright  sky  within  the  field  of  vision,  or  else  they  are  likely 
to  sit  in  such  a  position  that  their  faces  are  not  well  lighted.  The 
light  is  incident  upon  horizontal  surfaces  at  a  very  sharp  angle, 
and  there  is  only  one  good  position  for  writing,  or  two  good  posi- 
tions for  reading,  if  glare  from  the  window  or  the  paper  is  to  be 
avoided,  and  as  well  shadow  from  the  body  or  hand.  Practically 
the  only  way  of  bettering  these  conditions  which  has  been 
developed  so  far,  is  to  utilize  a  window  shade  to  protect  the  eye 
against  direct  light  from  the  sky,  and  this  is  done  of  course  at 
the  expense  of  the  illumination  of  the  room.  The  usual  direc- 
tion of  the  light  is  in  my  opinion  objectionable  both  from  the 
standpoint  of  utility  and  good  appearance  of  the  room.  The 
proper  utilization  of  daylight  for  interior  illumination  is  a  sub- 
ject of  which  the  study  has  not  yet  been  undertaken  seriously. 

There  is  one  quality,  however,  in  daylight,  whether  out-of- 
doors  or  indoors,  which  has  until  recently  been  lacking  in  our 
artificial  lighting — and  that  is,  ample  diffusion.  Interiors  are 
illuminated  as  a  rule  from  a  portion  of  the  sky,  the  light  source 
being  as  large  as  the  unobstructed  portion  of  the  window.  Out- 
of-doors,  even  in  brilliant  sunlight,  the  skylight  is  a  considerable 
factor  in  the  total  illumination.  Of  recent  years  more  attention 
has  been  given  to  this  quality  of  diffusion,  which  previously  had 
been  lacking  in  our  artificial  lighting.  Early  consideration  of 
lighting  principles  brought  realization  of  the  harm  which  exposed 
light  sources  work,  and  led  to  attempts  to  conceal  the  light  source. 
There  was  evolved,  among  other  systems,  that  of  cove  lighting. 
In  the  process  of  concealing  the  lamps  and  permitting  the  light 
to  fall  upon  a  white  surface,  from  which  a  part  of  it  was  re- 
flected into  the  room,  the  light  was  thoroughly  diffused.     This 


114  TRANSACTIONS  I.  E.  S. — PART  II 

system  of  lighting  is  more  notable  in  regard  to  the  success  with 
which  it  concealed  the  light  sources  and  diffused  the  light,  than 
in  other  respects.  Historically,  it  is  notable  for  the  evidence 
which  it  affords  of  growing  appreciation  of  some  of  the  principles 
of  good  lighting  which  are  now  considered  to  be  thoroughly 
established.  The  trouble  with  cove  lighting  as  usually  applied,  is 
that  control  of  the  direction  of  the  light  is  lost,  and  that  the  flux 
which  is  permitted  to  escape  from  the  cove  is  diffused  promiscu- 
ously throughout  the  room,  producing  a  flat  and  characterless 
effect.  Only  a  small  portion  of  the  flux  is  directed  where  it  is 
most  wanted,  while  perhaps  an  equal  portion  is  permitted  to  fall 
upon  surfaces  where  it  is  not  desired  in  such  quantities.  The 
system  has  not  been  largely  applied,  it  being  found  possible  to 
realize  its  advantages  by  other  methods  which  are  free  from  some 
of  its  disadvantages. 

More  recently  another  system  of  indirect  lighting  has  been 
developed,  in  which  central  fixtures  are  employed  to  conceal  the 
lamp  from  view  and  direct  much  of  its  light  to  the  ceiling,  from 
which  surface  it  is  diffused  downward.  More  engineering  study 
has  been  devoted  to  this  system  of  lighting,  and  in  consequence 
its  possibilities  have  been  more  largely  realized,  than  were  the 
possibilities  of  cove  lighting.  This  system  of  indirect  lighting 
has  been  widely  exploited,  and  has  given  considerable  satisfaction 
in  a  wide  variety  of  installations. 

Direct  lighting,  in  which  the  great  bulk  of  the  light  utilized 
comes  directly  from  the  light  source,  had  been  abused  with 
detrimental  results.  Particularly  was  it  lacking  in  diffusion. 
Indirect  lighting  is  the  other  extreme,  possessing  in  a  high  degree 
the  element  of  diffusion  which  is  so  often  lacking  in  direct  light- 
ing systems.  The  rapid  growth  of  indirect  lighting  is  the  mani- 
festation of  a  protest  against  abuse  of  direct  lighting.  Its  effect 
has  been  to  introduce  into  direct  lighting  practise  a  considerable 
general  improvement,  which  has  corrected,  or  decreased  some 
of  the  evils  of  direct  lighting.  And  too  much  credit  cannot  be 
given  to  the  exploiters  of  indirect  lighting  devices  for  the  bene- 
ficial influence  which  they  have  exerted  upon  our  lighting  prac- 
tice in  general. 

In  the  lighting  fixtures  here  shown  (Fig.  10),  the  lamp  in  the 
metal  bowl  is  backed  by  an  efficient  mirrored  reflector,  which 


m 


Viewed  from  direction  in  which        Viewed  from  direction  in  which 
glare  is  not  apparent.  glare  is  apparent. 

Fig.  6. — Showing  glare  due  to  specular  reflection  from  glossy  surface. 


Good 
reading  posture 


Undesirable 
reading  posture. 

Fig.  7. — Showing  manner  in  which  a  table  lamp  may  be  used. 


Very 
undesirable  reading  posture. 


F'g-  8.— Illumination  from  wall  brackets  alone. 


Fig.  9.— Daylight  illumination. 


Room  3. 


Room  2. 
Fig.  11. — Semi-indirect  lighting 


Room  1. 


*•» 

IBs         v': 

*^  -•-       .  .„ a    - .  1 

_ ; "j-Cy, 

,  ,..                       -■■'■-—< 

Indirect.  Semi-indirect.  Direct. 

Fig.  12. — Three  common  methods  of  illumination. 


Room  3.  Room  2.  Room  1. 

Fig.  13. — Suggesting  decorative  or  ornamental  lighting  units. 


MILLAR:      ILLUMINATION  OF  INTERIORS  11$ 

directs  its  light  toward  the  ceiling.  The  rooms  have  been 
equipped  with  three  ceilings — one  is  white,  and  has  about  as 
high  a  reflecting  coefficient  as  is  likely  to  be  found  in  practise. 
Another  is  cream  colored,  and  reflects  a  smaller  proportion  of 
the  light.  A  third  is  dark  cream,  approaching  a  tan,  and  reflects 
still  less  of  the  light.  This  latter  is  about  as  dark  as  one  might 
expect  to  find  employed  in  an  indirect  lighting  system,  where  any 
attention  is  paid  to  efficiency.  Indirect  lighting  is  so  largely 
dependent  upon  the  reflecting  qualities  of  the  ceiling,  that  the 
statistics  of  the  illumination  intensities  in  these  rooms  are  of 
interest.  The  horizontal  illumination  intensity  on  the  table  plane 
averages  for  the  three  ceilings  : 

White  ceiling 100% 

Light  cream  ceiling* 87  Jo 

Dark  cream  ceiling 58  % 

showing  a  reduced  efficiency  of  42  per  cent,  due  to  the  inferior 
reflecting  qualities  of  the  darker  ceiling. 

Following  closely  upon  the  development  of  the  indirect  light- 
ing system,  come  systems  classed  inaccurately  as  semi-indirect 
lighting  units,  in  which  part  of  the  light  is  reflected  from  the 
ceiling,  as  in  the  indirect  system,  while  part  of  it  comes  directly 
from  the  translucent  bowl  surrounding  the  light  source.  It  is 
obvious  of  course  that  with  any  translucent  lighting  auxiliary 
employed  in  a  direct  lighting  system,  some  of  the  light  which 
reaches  the  ceiling  and  walls,  is  reflected  downward,  and  that 
the  system  is  thus  a  semi-indirect  system.  Those  units  which 
are  classed  as  semi-indirect  units  at  the  present  time  are,  however, 
units  designed  especially  with  a  view  to  directing  a  considerable 
proportion  of  the  light  toward  the  ceiling  The  most  desirable 
combination  of  direct  and  indirect  light  for  general  purposes 
served  by  such  units,  is  to-day  a  subject  of  discussion.  Views  of 
illuminating  engineers  vary  in  this  matter.  All  kinds  of  relations 
between  these  two  components  are  to  be  found  represented  by 
outfits  now  available  in  the  open  market.  These  range  from 
equipments  in  which  the  transmitted  light  is  so  small  a  proportion 

*  It  was  discovered  too  late  for  correction  that  the  light  cream  ceiling-  has  diffusing 
qualities  so  unlike  the  dark  cream  ceiling  that  in  spite  of  reasonably  typical  intensities 
on  the  table  plane,  the  appearance  as  viewed  from  without  the  rooms  is  not  consistent 
with  the  intensity  figures  shown  ;  thus,  the  ceiling  in  Room  2,  when  viewed  from  the 
table,  is  much  lighter  than  the  ceiling  in  Room  3  though  it  does  not  so  appear  in  the 
figures. 


Il6  TRANSACTIONS  I.  E.  S. — PART  II 

of  the  total  as  to  make  it  apparent  that  the  purpose  to  be  served 
by  the  direct  component  is  chiefly  one  of  decoration,  to  those  in 
which  the  direct  component  is  so  large  as  to  make  evident  an 
intention  to  increase  the  efficiency  considerably  by  restricting  the 
amount  of  light  which  is  subjected  to  the  inherent  ceiling  loss. 

In  the  photograph  (Fig.  n)  three  semi-indirect  lighting  fix- 
tures are  shown.  In  room  No.  I,  a  direct  lighting  reflector  is 
inverted.  In  room  No.  3,  a  bowl,  not  intended  for  this  purpose, 
is  employed.  The  design  of  its  surfaces  is  not  well  adapted  to 
this  purpose,  and  it  is  therefore  not  so  efficient  as  it  might 
otherwise  be  made.  In  room  No.  2  a  hemisphere  is  utilized, 
illustrating  semi-indirect  lighting  in  the  simplest  of  its  charac- 
teristic forms. 

It  is  a  matter  for  gratification  that  illuminating  engineers 
to-day  have  such  an  excellent  choice  as  that  afforded  by  the  wide 
range  of  available  equipments  for  direct,  indirect  and  semi- 
indirect  lighting  systems.  Each  has  its  merits,  each  its  demerits. 
In  some  installations,  one  type  is  preferable,  in  other  installations, 
some  other  type  may  produce  most  desirable  results.  The  good 
qualities  which  characterize  each  are  coming  to  be  incorporated, 
as  far  as  practicable,  in  the  others,  and  it  may  be  noted  that  the 
more  vigorously  each  system  is  exploited,  the  more  beneficial 
upon  lighting  practise  in  general  will  the  result  be.  With  a 
direct  lighting  system,  it  is  a  simple  matter  to  direct  a  relatively 
large  percentage  of  the  light  downward  upon  say  the  table  plane, 
but  it  is  a  difficult  matter  to  so  dispose  the  lamps  and  to  so 
equip  them  that  the  installation  will  be  free  from  troubles  due 
to  glare  and  shadow.  With  an  indirect  lighting  system  it  is 
relatively  a  simple  matter  to  avoid  deleterious  effects  due  to 
glare  and  shadow,  but  it  is  very  difficult  to  direct  a  satisfactorily 
large  percentage  of  the  light  upon  the  table  plane.  Where  absence 
of  glare  and  shadow  is  a  consideration  of  paramount  importance, 
an  indirect  or  a  semi-indirect  lighting  system  may  often  be 
preferable,  in  spite  of  the  necessity  for  somewhat  greater  expen- 
diture in  energy.  Where  these  considerations  are  not  so  im- 
portant, or  where  economy  of  operation  is  the  prime  considera- 
tion, a  direct  lighting  system  may  prove  preferable.  In  any  case, 
the  adroitness  of  the  illuminating  engineer  may  exhibit  itself  in 


miliar:     illumination  of  interiors  117 

securing  the  best  balance  between  economy  on  the  one  hand 
and  absence  of  glare  and  shadow  on  the  other.  As  to  the  appear- 
ance of  the  installations,  there  may  be  all  kinds  of  diverse  views, 
and  we  must  remember  that  there  is  no  disputing  taste.  Obvi- 
ously it  is  difficult  to  discuss  those  phases  of  the  question  when 
dealing  with  the  subject  in  a  general  way. 

The  three  modern  systems  of  lighting  are  represented 
in  Fig.  12.  In  room  No.  1,  the  direct  lighting  unit 
transmits  sufficient  light  to  make  the  walls  pleasantly,  but  not 
objectionally,  bright,  while  directing  much  of  the  light  to  the 
table  plane.  In  room  No.  2,  the  semi-indirect  unit  illuminates 
the  card  by  light  direct  from  the  bowl  and  by  light  from  the 
ceiling  and  walls  in  something  like  the  proportions  of  3  to  I.  The 
relative  direct  and  indirect  components  upon  the  table  plane  are 
of  the  order  of  iy2  to  1.  In  room  No.  3  all  of  the  light  is 
diffused  from  the  ceiling.  The  ceiling  is  the  brightest  surface 
within  view,  the  lamp  being  entirely  concealed.  The  illumination 
is  very  soft  and  uniform. 

Comparing  the  two  end  rooms,  it  will  be  noted  that  in  room 
No.  1,  the  vertical  stripes  in  the  wall-paper  may  be  seen  standing 
out  clear  and  sharp.  The  character  of  the  pattern  is  evident.  In 
room  No.  3  these  stripes  are  seen  somewhat  less  distinctly.  This 
is  due  to  the  lower  intensity  of  light  on  the  wall.  Still  more  im- 
portant, however,  as  a  factor,  is  the  downward  direction  of  the 
light  from  the  ceiling.  Viewed  from  the  table,  these  stripes  stand 
out  distinctly  as  the  angle  and  direction  are  then  such  as  to  be 
within  the  zone  of  strong  specular  reflection  from  the  wall-paper. 
Viewed  as  in  the  photograph,  these  stripes  can  hardly  be  dis- 
cerned except  on  the  upper  part  of  the  walls  near  the  border. 
The  paper  loses  its  character.  This  is  an  excellent  illustration  of 
the  importance  of  securing  proper  direction  from  the  major  part 
of  the  light,  although  it  should  not  be  taken  as  an  indication  that 
the  direction  is  wrong  in  this  installation  because  it  must  be 
remembered  that  the  effect  would  be  minimized  if  the  wall-paper 
were  viewed  from  within  the  room,  instead  of  from  without. 

With  the  conditions  as  established  (and  it  is  not  claimed  that 
they  are  more  than  suggestive  of  typical  conditions)  the  card 
illuminations  are  as  follows : 


Il8  TRANSACTIONS  I.  E.  S. — PART  II 

Room  No.  I 220% 

Room  No.  2 100  Jo 

Room  No.  3 42  f0 

It  must  be  remembered  however,  that  the  direct  lighting  unit  in 
this  case  is  favored,  because  the  card  is  immediately  beneath  it 
at  the  point  of  highest  intensity.  For  purposes  of  reading,  as  an 
example,  it  is  difficult  to  judge  from  these  figures  as  to  the  rela- 
tive useful  light.  In  the  first  place,  questions  of  diffusion  may 
result  in  establishing  demands  for  higher  intensities  in  one  system 
than  in  another.  This  is  one  of  the  questions  which  is  being 
very  generally  investigated  at  the  present  time,  and  in  such  a  re- 
view as  this,  its  discussion  has  no  place.  Dealing  solely  with  the 
question  of  distribution,  it  may  be  noted  that  most  reading  would 
be  likely  to  be  done  near  the  center  of  the  room  and  that  there- 
fore the  direct  lighting  system  should  receive  some  of  the  advan- 
tage in  rating  which  the  high  intensity  of  the  card  immediately 
beneath  the  unit  would  appear  to  give  it.  The  relative  higher 
intensities  in  the  corners  of  the  room  with  the  indirect,  and  to  a 
lesser  degree  with  the  semi-indirect  fixture,  are  not  of  much 
advantage  from  a  practical  standpoint.  In  this  particular  in- 
stallation, with  the  same  flux  produced  by  the  lamps  in  each 
type  of  lighting,  the  average  horizontal  intensities  are  relatively: 

Direct  lighting 1 .61  foot-candles 

Semi-indirect  lighting 1.33  foot-candles 

Indirect  lighting 0.91  foot-candles 

It  is  generally  believed  that  with  conditions  suitable  for  each 
system  of  lighting,  the  direct  lighting  system  will  deliver  about 
twice  as  much  light  upon  the  table  plane  as  does  the  indirect 
lighting  system,  while  each  will  illuminate  the  walls  moderately. 

The  decorative  feature  has  kept  pace  with  developments  in  the 
other  branches  of  the  art.  Lighting  auxiliaries  are  consistently 
being  improved  in  appearance,  as  well  as  in  other  features  of 
effectiveness.  Efficiency  of  reflection,  the  necessary  degree  of 
diffusion,  and  the  proper  direction  of  light  are  being  achieved  more 
and  more  completely  as  experience  becomes  greater.  In  good 
taste  and  other  qualities  that  make  for  pleasing  effects,  constant 
advances  are  being  made  also. 

The  older  lines  of  lighting  glassware,  including  the  prismatic, 
have  been  modified  so  as  to  render  them  more  pleasing  in  appear- 


millar:     illumination  of  interiors  119 

ance  while  the  addition  of  a  number  of  new  lines  of  phosphate 
and  other  glass  affords  the  user  a  number  of  alternatives  in  the 
way  of  glassware  equipment  suitable  for  use  with  any  given  type 
of  fixture  in  any  ordinary  installation. 

In  Fig.  13,  may  be  seen  illustrations  of  some  of  the  more 
decorative  types  of  fixtures  and  glassware  now  available  in  stand- 
ard types.  Whatever  the  character  of  the  installation  may  be,  it 
is  more  than  likely  that  unless  it  is  extraordinary,  some  fixture 
and  some  kind  of  glassware  may  be  obtained  which  may  be  used 
in  the  installation  with  fair  satisfaction.  Unless  installations  are 
considered  which  are  so  unusual  as  to  demand  the  design  of 
special  lighting  equipments,  those  now  obtainable  must  be  con- 
sidered to  afford  a  very  satisfactory  range  of  selection. 

The  foregoing  demonstrations  must  be  taken  with  some  quali- 
fications. They  do  not  pretend  in  all  cases  to  be  typical  of  any 
particular  class  of  installation.  Time  has  not  permitted  a  thor- 
ough discussion  of  the  characteristic  qualities  of  any  one  of  them. 
Appearances  have  been  different  from  those  which  would  have 
presented  themselves  had  the  rooms  been  observed  from  within. 

The  one  thing  which  seems  to  me  to  have  an  immediate  bearing 
from  the  central  station  standpoint,  and  which  I  hope  is  re- 
cognized is  that  if  artificial  lighting  is  to  be  made  thor- 
oughly good  and  satisfactory,  it  is  necessary  to  thoroughly 
diffuse  and  otherwise  modify  the  light  which  is  produced  by  the 
lamps.  This  cannot  be  accomplished  without  considerable  loss 
of  light,  and  therefore  entails  greater  consumption  of  electrical 
energy.  Thus  immediate  commercial  advantage  goes  hand  in 
hand  with  good  business  policy  and  with  altruism,  when  the  cen- 
tral station  spreads  the  gospel  of  good  lighting  among  its 
customers. 


120  TRANSACTIONS  I.  E.  S. — PART  II 

DISCUSSION  AT  A  MEETING  OF  THE  NEW  ENGLAND 
SECTION,  MARCH   13,   1913. 


Mr.  A.  E.  Jossexyn  :  I  think  Mr.  Millar's  demonstration  has 
brought  before  us  strongly  the  conclusion  which  probably  many 
of  us  had  reached  in  investigating  complaints  of  poor  lighting, 
viz.,  it  is  not  so  much  a  question  of  the  light  as  it  is  of  the 
lighting,  the  kind  of  glassware,  fittings  or  the  radiation  of  the 
light  itself.  It  is  not  so  much  a  question  of  the  service  supplied 
by  the  lighting  company.  In  my  experience,  covering  several 
years,  we  have  many  times  been  called  upon  to  investigate  what 
bur  customers  claimed  to  be  poor  light  and  found,  as  a  matter  of 
fact,  that  the  cause  was  poor  lighting  rather  than  poor  light; 
that  is  to  say  the  distribution  of  the  light  has  been  the  cause  of 
the  dissatisfaction.  In  many  cases  this  has  been  due  to  the 
location  of  the  fixture  or,  if  that  has  been  located  properly,  it 
has  been  due  to  the  fact  that  the  glassware  was  selected  more  for 
decorative  than  for  lighting  purposes.  I  believe  Mr.  Millar's 
paper  has  brought  out  the  fact  that  it  is  not  so  much  the  light  as 
it  is  the  lighting. 

Mr.  R.  C.  Ware  :  I  think  there  is  another  point  that  is  very 
strongly  brought  out  along  the  lines  just  mentioned.  We  get 
complaints  sometimes  of  poor  light  which  is  not  only  due  to 
poor  distribution  but  to  the  fact  that  the  customer  has  insisted 
on  using  bare  lamps.  I  think  this  very  valuable  demonstration 
of  Mr.  Millar's  has  brought  out  very  clearly  the  absolute  necessity 
of  protecting  the  eyes ;  and  that  bare  lamps  mean  the  stopping 
down  of  the  pupil  of  the  eye  to  such  an  extent  that  the  user 
does  not  get  the  full  benefit  of  the  light  actually  given  off.  This 
certainly  ought  to  help  us  to  talk  intelligently  and  forcibly  to 
customers  who  complain  that  they  do  not  get  results.  We  ought 
to  be  able  to  show  them  now  why  they  do  not,  and  to  help  them 
on  to  the  road  so  that  they  may  get  what  they  are  after. 

Mr.  J.  W.  CowlEs:  I  judge  that  others  are  affected  in  the 
same  way  that  I  am  to-night,  in  being  somewhat  over-awed  by 
this  paper  in  its  many  opportunities  for  discussion.  What  we 
have  seen  and  heard  has  been  brought  out  in  such  rapid  succes- 
sion that  for  my  part  I  am  quite  bewildered  in  knowing  just 
where  to  enter  into  the  subject  from  a  discussion  standpoint. 


ILLUMINATION   OF  INTERIORS  121 

There  is  so  much  that  we  have  seen, — so  many  points  have  been 
brought  out  in  a  vivid  and  interesting  manner, — that  a  con- 
siderable amount  of  careful  thought  is  required  for  the  proper 
digestion  of  the  material  offered.  I  have  not  as  yet  had  an 
opportunity  to  read  this  paper  in  print,  but  my  firm  resolve 
to-night  is  to  sit  down  with  this  paper  in  the  quiet  of  my  room, 
go  through  it  carefully,  and  gain  from  it  the  profit  that  must 
come  from  the  perusal  and  more  gradual  study  of  it. 


Prof.  George  A.  HoadeEy  :  I  think  one  thing  in  the  indirect 
lighting  system  should  be  taken  into  account,  that  is,  the  eye 
accommodates  itself  to  the  light  in  the  room;  you  get  the  im- 
pression when  first  coming  into  the  room,  that  the  illumination 
is  insufficient,  but  after  having  been  in  the  room  a  short  time, 
the  illumination  becomes  sufficient.  I  might  cite  an  example  of 
the  lighting  of  a  dining-room  in  which  an  inverted  cone  shade 
was  used  which  gave  a  spotty  light  on  the  table,  and  after 
putting  a  piece  of  ground  glass  across  the  bottom  of  the  angle 
shade,  the  illumination  became  satisfactory. 


122  TRANSACTIONS  I.  E.  S. PART  II 

DISCUSSION  AT  JOINT  MEETING  OF  I.  E.   S.   NEW 
YORK  SECTION  AND  NEW  YORK  COMPANIES' 
SECTION  OF  THE  NATIONAL  ELEC- 
TRIC LIGHT  ASSOCIATION, 
NOVEMBER  18,  1912. 


Dr.  Herbert  E.  Ives:  The  demonstration  which  Mr.  Millar 
has  given  us  this  evening  is,  in  my  opinion,  one  of  the  most 
instructive  ever  given  on  the  subject  of  illumination.  We  all 
realize  from  his  use  of  dolls  and  toy  furniture  that  this  talk  was 
meant  to  be  of  a  kindergarten  nature.  Nevertheless  I  feel  sure 
that  I  am  speaking  for  all  present,  even  those  who  in  the  words 
of  the  chairman  have  spent  a  lifetime  studying  illumination,  when 
I  say  that  we  have  learned  a  very  great  deal  from  his  clear 
presentation  and  admirable  demonstration.  I  wish  to  make  no 
criticism  of  this,  for  it  deserves  nothing  but  praise. 

Of  course  no  one  studying  such  a  comparatively  new  subject 
as  lighting  will  agree  with  everyone  else.  I  would  like  to  take 
this  opportunity  to  emphasize  a  point  which  Mr.  Millar  could 
not,  in  the  time  at  his  disposal,  treat.  I  want  to  call  attention  to 
the  fact  that  in  all  these  demonstration  booths  the  light  source  is 
a  centrally  overhead  fixture.  We  are  so  accustomed  to  such  a 
system  of  lighting  that  I  think  we  are  apt  to  overlook  the  fact 
that  it  is  not  the  only  possible  method  and  perhaps  not  even  the 
best.  The  other  evening  there  was  a  paper  presented  at  the 
Philadelphia  Section  of  the  Illuminating  Engineering  Society  on 
indirect  lighting,  and  one  remark  made  by  the  speaker  was  to 
me  very  suggestive.  He  said  that  all  of  the  fixtures  used  were 
placed  on  the  outlets  which  had  been  "planned  by  the  contractor." 
It  is  a  question  in  my  mind  whether  we  have  not  reached  the 
point  where  we  must  go  beyond  the  contractor  and  his  ideas. 
For  instance,  may  it  not  be  possible  that  the  light  of  the  future 
will  be  from  the  side  rather  than  from  overhead?  The  lighting 
of  a  room  by  daylight  is  from  the  side  and  is  generally  considered 
pretty  satisfactory.  It  differs,  too,  from  most  artificial  lighting 
in  the  size  of  the  light  sources. 

Mr.  Millar  has  followed  the  usual  classification  of  direct,  indi- 
rect and  semi-indirect  lighting.  To  my  mind,  however,  the 
proper  classification  is  on  the  basis  of  the  size  of  the  sources. 


ILLUMINATION  OF  INTERIORS  123 

We  have  been  accustomed  to  small  light  sources  which  are 
necessarily  of  high  intrinsic  brilliancy,  and  we  are  now  working 
towards  larger  sources  of  lower  intrinsic  brilliancy,  whether  it 
be  by  the  use  of  translucent  diffusing  media  or  by  diffusive 
reflecting  material  on  the  ceilings.  In  the  illumination  of  most 
rooms  by  daylight  we  have  a  very  large  light  source,  namely,  the 
sky,  which  at  the  same  time  is  usually  not  visible  to  a  person  in 
the  room  who  looks  out  at  the  neighboring  houses  or  at  the 
lower  portions  of  the  landscape.  The  net  result  is  illumination 
from  a  large  invisible  light  source  from  which  the  general  direction 
of  illumination  is  at  the  side.  I  hope  to  see  experiments  made 
with  a  view  to  meeting  these  conditions  by  artificial  light.  I 
feel  sure  they  will  be  instructive  and  they  may  lead  to  some 
satisfactory  systems  of  artificial  lighting. 

Right  in  line  with  Mr.  Millar's  concluding  remarks  I  may  say 
that  according  to  all  present  indications  if  we  do  copy  daylight 
illumination  in  the  manner  I  have  suggested  it  will  mean  the  use 
of  an  enormously  greater  amount  of  electrical  energy  or  gas. 

Dr.  C.  H.  Sharp:  I  feel  very  much  as  Dr.  Ives  that  there  is 
much  to  be  said  in  the  way  of  praise  about  what  we  have  seen 
here  to-night.  I  have  known  for  some  time  that  Mr.  Millar  had 
in  preparation  a  demonstration  of  light  in  miniature,  and  I  have 
been  wondering,  without  knowing  anything  about  it,  what  he 
could  make  out  of  it.  How  could  he  show  to  us  lighting  effects  in 
rooms  which  we  are  not  inside  of  but  which  we  are  merely  allowed 
to  look  into  and  give  us  any  adequate  idea  of  what  is  really  the 
effect  inside  those  rooms  ?  My  questions  have  all  been  answered. 
I  can  say  that  he  has  been  able  to  show  us  a  great  deal  regarding 
practical  conditions  of  illumination. 

I  agree  with  Dr.  Ives  regarding  daylight  illumination. 
I  think  it  is  pretty  good,  and  one  reason  why  it  is 
good  is  because  there  is  plenty  of  it,  and  if  there  is 
not  plenty  of  it  it  is  not  good.  Under  the  usual  New  York  con- 
ditions we  are  not  often  much  affected  with  the  glare  from  the  sky. 
We  do  not  see  the  sky,  but  we  do  get  a  thoroughly  diffused  light 
in  the  room  which  is  sufficient  in  amount  but  which  is  directed. 
Xow  I  think  that  it  might  be  possible  to  work  out  something  in 
artificial  illumination  to  simulate  daylight  illumination.  Imagine 
3 


124  TRANSACTIONS  I.  E.  S. — PART  II 

you  have  around  each  window  of  the  room,  say,  a  trough  re- 
flector with  a  lot  of  lamps  so  directed  that  if  they  were  turned  on 
they  would  throw  light  out  of  the  room.  Now  when  night  comes 
on  and  light  ceases  to  stream  in  through  the  windows,  we  draw 
down  over  the  window  a  very  white  window  shade,  opaque,  and 
with  a  very  good  white  surface,  and  then  turn  on  the  lights.  Then 
the  lamps  will  throw  a  strong  illumination  on  the  window  shade 
and  which  will  throw  it  back  into  the  room  directed  and  diffused 
and  distributed  similarly  to  daylight.  There  should  be  less  glare 
than  there  is  in  daylight.  This  would  have  another  advantage  in 
that  if  the  office  furniture  and  the  arrangements  of  the  room  in 
general  were  made  so  as  to  be  most  advantageous  for  daylight 
illumination,  they  would  also  be  the  most  advantageous  arrange- 
ments for  the  night  time.  I  have  not  made  any  computations  on 
this  as  to  the  number  of  watts  per  cubic  foot  it  would  require,  but 
I  have  no  doubt  that  the  plan  would  be  practicable  in  certain  cases 
even  if  the  efficiency  were  not  very  high.  It  could  hardly  be  more 
extravagant  than  other  indirect  systems. 

Mr.  L.  B.  Marks  :  As  I  heard  Mr.  Millar  talking  this  even- 
ing I  recall  a  letter  which  I  received  about  two  years  ago  from 
Dr.  Ives  in  which  he  suggested  that  I  make  similar  demonstra- 
tions at  the  Johns  Hopkins  University  in  Baltimore  in  connection 
with  my  lectures  on  principles  and  designs  of  interior  illumina- 
tion; but  I  was  afraid  to  try  it.  I  did  not  have  the  nerve  Mr. 
Millar  has.  As  I  sat  here  this  evening  and  heard  his  lecture  and 
viewed  the  demonstrations  of  the  principles  discussed  by  me  at 
Baltimore,  I  felt  convinced  that  anybody  who  had  the  good 
fortune  to  see  the  demonstration  which  Mr.  Millar  has  prepared 
would  receive  an  object  lesson  in  illumination  which  I  believe 
would  be  far  better  than  any  theoretical  discussion  of  the  subject. 

With  regard  to  the  question  that  was  brought  up  this  evening 
of  producing  a  distribution  and  direction  of  light  similar  to  that 
of  daylight,  I  have  a  somewhat  different  view  than  that  expressed 
by  the  other  speakers.  It  seems  to  me  that  it  is  desirable  to  have 
a  change  at  night.  We  all  want  a  little  variety.  We  do  not  want 
to  eat  the  same  dish  at  every  meal,  breakfast,  lunch  and  dinner. 

For  daylight  illumination  we  must  depend  almost  exclusively 
on  side  windows.     It  is  a  fact  that  we  cannot  plan  the  best  day- 


ILLUMINATION  OF  INTERIORS  125 

light  illumination  because  of  the  physical  limitations  of  buildings. 
We  usually  have  a  number  of  floors  in  each  building  and  cannot 
get  a  desirable  distribution  of  light  in  all  of  them,  if  indeed  in 
any  of  them.  We  have  not  that  limitation  at  night.  Then  why 
not  avail  ourselves  of  the  broad  scope  of  application  we  have  in 
electric  and  gas  lighting?  If  you  discuss  this  matter  with  the 
architect  or  with  the  decorator  he  is  likely  to  tell  you  he  can 
obtain  much  more  pleasing  effects  at  night  by  artificial  light  than 
he  can  in  the  day  time,  because  he  has  the  ability  to  place  the 
light  sources  where  he  wants  them.  One  of  the  criticisms  which 
Mr.  Edward  Caldwell  the  distinguished  fixture  designer  and 
decorator  made  of  my  lectures  was  that  I  did  not  lay  sufficient 
stress  on  the  importance  of  changing  the  character  of  the 
illumination  at  night.  In  his  opinion  that  constituted  one  of  the 
important  things  that  an  illuminating  engineer  has  to  study  up — 
the  arrangements  of  various  lights  and  shades  and  the  production 
of  pleasing  effects  at  night,  not  possible  by  daylight. 

There  is  another  matter  which  I  want  to  speak  of  briefly, 
namely,  Mr.  Millar's  closing  remarks  with  reference  to  the 
greater  use  of  light.  In  the  I.  E.  S.  Primer  which  was  distributed 
this  evening  the  principles  underlying  nearly  all  of  Mr.  Millar's 
demonstrations  are  discussed  and  illustrated.  The  keynote  of 
this  Primer  is  good  diffusion  of  light.  What  does  good  diffusion 
of  light  mean  ?  It  usually  means  more  light  and  better  light.  I 
hold  that  it  is  up  to  the  illuminating  engineer  if  he  finds  it  neces- 
sary to  recommend  more  light  to  do  so  and  to  do  so  fearlessly, 
even  though  the  customer  is  told  he  must  pay  more  money  for  it. 
One  of  the  first  things  that  the  lighting  company's  representative 
will  be  up  against  when  the  public  calls  for  greater  diffusion  of 
light  is  the  complaint  that  the  lighting  bills  are  larger.  We  have 
seen  here  to-night  a  demonstration  which  will  convince  any  man — 
and  what  is  more  important  any  woman  of  the  household — that 
it  pays  well  to  use  more  current  or  gas  for  lighting  if  you  can 
get  better  results  in  diffusion  of  light. 

It  would  not  be  a  bad  idea  to  get  out  a  miniature  exhibition  set 
of  this  kind  that  the  salesman  or  company's  representative  could 
take  with  him  to  make  a  demonstration  to  the  housewife. 

Mr.  D.  McFarlan  Moore:  Mr.  Millar  has  laid  before  us 
a  wonderful  wealth  of  ideas.     We  can  hope   for  some  future 


126  TRANSACTIONS  I.  E.  S. — PART  II 

paper  to  carry  out  in  greater  detail  some  of  the  modifications  of 
his  general  scheme  that  suggest  themselves.  But  I  admire  the 
thoroughness  with  which  he  has  worked  out  the  equipment  details 
of  his  three  demonstration  rooms. 

In  these  days  we  hear  a  great  deal  about  direct,  semi-direct  and 
indirect  lighting.  Still  better  diffusion  is  continually  desired. 
Indirect  lighting  virtually  consists  in  changing  a  point  source  to 
the  entire  ceiling  as  the  source  of  light. 

I  have  been  introduced  as  the  inventor  of  the  Moore  light.  We 
should  not  forget  that  one  of  the  solutions  of  this  problem  of 
better  diffusion  is  to  directly  increase  the  area  of  the  light  source 
itself,  abandoning  the  intense  light  of  bulbs  for  the  soft  light  of 
long  tubes,  thereby  also  avoiding  the  necessity  for  any  form  of 
reflecting  or  diffusing  or  softening  glassware,  but  using  the  light 
directly  at  the  intensity  at  which  it  is  generated. 

Dr.  A.  S.  McAllister  :  I  have  nothing  to  say  in  addition  to 
the  remarks,  except  that  I  wish  to  compliment  Mr.  Millar  on  his 
excellent  demonstration,  the  thoroughness  of  it,  the  accuracy  and 
fineness  of  the  work,  etc.  The  subject  of  light  from  the  aspect 
which  Dr.  Ives  spoke  of  is  one  to  which  I  have  given  some  little 
thought.  I  do  not  quite  agree  with  him  that  we  want  the  light  on 
the  side.  In  order  to  get  it  on  the  side  we  must  place  it  in  the 
line  of  view.  Now  I  do  not  care  how  much  it  diffuses,  we  must 
have  a  little  of  that,  and  that  to  my  mind  is  not  objectionable.  I 
am  inclined  to  think  that  light  above  is  oftentimes  more 
advantageous. 

Mr.  C.  A.  Ljttlefteld  :  I  wish  to  express  to  Mr.  Millar 
personally  and  on  behalf  of  the  section  my  appreciation 
of  the  wonderful  demonstration  he  has  given  us  this 
evening.  Its  chief  advantage  is  that  it  is  so  complete 
and  comprehensive,  yet  so  simple  and  devoid  of  unin- 
teresting technical  detail.  To  my  mind  one  of  the  most  success- 
ful things  the  Illuminating  Engineering  Society  has  done,  from 
a  popular  standpoint,  is  the  publication  of  the  Primer.  If  in 
some  manner  this  lecture  and  the  Primer  could  be  made  integral 
parts  and  be  presented  in  various  parts  of  the  city  and  country 
a  vast  amount  of  good  could  be  accomplished.  We  all  know 
that  very  little  is  known  of  correct  lighting  principles,  and  even 


ILLUMINATION  OF  INTERIORS  12"J 

in  my  own  home  I  know  that  lighting  conditions  are  not  as  they 
should  be,  but  could  this  lecture  be  given  before  popular  audi- 
ences, in  churches,  lecture  halls,  etc.,  it  would  get  before  the 
public  generally  a  knowledge  of  correct  lighting  fundamentals 
that  would  do  an  enormous  amount  of  good.  It  is  too  good  to 
be  kept  merely  for  technical  societies  and  meetings,  and  I  trust 
that  some  way  may  be  found  whereby  this  plan  can  be  carried 
out.  I  feel  that  this  must  have  been  one  of  the  motives  that 
animated  Mr.  Millar  to  give  such  an  enormous  amount  of  his 
valuable  time  to  prepare  this  lecture.  Mr.  Millar  is  too  well 
and  favorably  known  among  the  profession  to  need  any  eulogy 
from  me,  but  a  public  delivery  of  this  lecture  would  much 
enhance  his  reputation  and  at  the  same  time  do  a  great  amount 
of  good. 

Mr.  Norman  Macbeth  :  There  is  very  little  I  can  add.  I  con- 
sider the  demonstration  Mr.  Millar  has  made  here  to-night  one  of 
the  finest  things  I  have  seen  in  a  long  time.  His  methods  of  show- 
ing the  wonderful  detail  work  of  the  various  sections  of  these 
rooms,  the  smoothness  of  the  stage  work,  and  his  lecture  itself 
are  remarkable.  I  believe  a  great  many  people  here  have  an 
entirely  different  idea  of  the  lighting  question  after  seeing  Mr. 
Millar's  demonstrations. 

Mr.  A.  J.  Marshall  :  I  wish  to  express  my  congratulations 
to  Mr.  Millar  on  his  most  clever,  instructive,  and  entertaining 
demonstration,  which  I  feel  privileged  to  have  been  able  to 
witness. 

I  am  a  very  great  believer  in  catering  to  the  brain  through 
the  eye  in  educational  work,  and  consequently  am  the  better 
able  to  appreciate  the  character  of  Mr.  Millar's  elaborate  and 
painstaking  experiments. 

It  would  seem  unpardonable  to  criticise  the  effects  that  we 
have  witnessed  this  evening  on  account  of  their  general  all 
around  value.  However,  I  would  like  to  state  that  none  of  the 
interiors  which  we  have  seen,  to  me  represent  good  practise  in 
the  lighting  of  living  rooms.  As  a  matter  of  fact,  I  certainly 
would  not  employ  any  that  I  have  seen.  Most  of  the  effects 
shown  are  those  associated  with  the  past  or  with  what  might 
be  termed  inappropriate  "mechanical"  lighting  equipment,  which 


128  TRANSACTIONS  I.  E.  S. — PART  II 

certainly  does  not  represent  good  practise,  or  which  has  the  sup- 
port of  architects,  fixture  houses,  up-to-date  central  stations,  gas 
companies,  the  public,  etc.  In  fact,  it  is  just  the  equipment  as 
shown  that  advanced  workers  are  eliminating. 

This,  however,,  should  not  be  considered  as  being  uncom- 
plimentary to  Mr.  Millar  in  any  way,  shape,  or  form.  The  criti- 
cism is  directed  simply  to  equipment  which  is  to-day  not  con- 
sidered good  practise. 

The  interesting  phase  of  the  whole  matter  is  that  something 
has  been  done,  and  Mr.  Millar  should  be  complimented  upon  his 
success  in  doing  that  something  exceedingly  well. 

Mr.  A.  W.  Stark:  I  am  very  glad  to  tell  you  that  I  was 
very  much  impressed  with  Mr.  Millar's  lecture.  I  think,  how- 
ever, that  Mr.  Millar  omitted  a  very  important  subject  in  con- 
nection with  his  lecture,  and  that  was  the  quality  of  the  light. 
I  agree  with  Mr.  Marks,  who  spoke  a  moment  ago,  that  we  need 
a  variety  at  night.  If  it  is  possible  to  distribute  light  in  a  better 
manner  than  comes  through  window  openings  from  different 
floors  of  the  buildings  that  we  have  to  occupy  through  the  day, 
such  an  arrangement  is  much  more  desirable.  However,  the 
quality  of  the  light  is  vitally  important.  In  looking  around  I 
think  it  safe  to  say  that  25  per  cent,  of  those  here  are  wearing 
glasses.  Why?  I  think  that  if  we  had  somebody  present  who 
knows  more  of  the  subject  than  I  do  he  would  probably  say  they 
have  misused  their  eyes.  I  think  that  fact  is  generally  due  to 
not  only  misusing  the  eyes,  but  that  the  quality  of  light  was  not 
best  for  the  eyes.  Daylight  will  not  affect  as  seriously,  if 
properly  used,  the  eyes,  as  any  artificial  light.  I  would  have 
been  very  much  more  pleased  with  what  Mr.  Millar  had  to  say 
if  he  had  given  us  a  demonstration  of  what  gas  light  would  do. 

Mr.  C.  B.  Graves  :  The  question  of  indirect  lighting  and  low 
ceilings  probably  enters  more  into  residential  work  than  in  any 
other  work.  It  is  simply  a  matter  of  distribution,  and  in  that 
case  the  fixtures  must  be  suspended  within  a  distance  of  some 
14  or  18  inches  from  the  ceiling.  Very  good  results  are  obtained 
from  low  ceilings,  as  it  simply  means  a  matter  of  a  larger 
number  of  fixtures  for  the  area  to  be  covered.  I  think  that  where 
there  is  room  for  clearance,  that  is,  where  a  ceiling  of  &l/2  to 
9  feet  high  is  available,  that  indirect  lighting  can  be  worked  in 


ILLUMINATION  OF  INTERIORS  I2Q, 

very  successfully,  and  in  fact  it  has  been  done  in  a  great  many 
cases. 

Mr.  Preston  S.  Millar  (In  reply)  :  I  could  myself  also 
mention  a  few  things  that  have  been  overlooked  in  these  demon- 
strations. We  have  the  necessary  equipment.  We  have  a  de- 
finite schedule  for  about  40  demonstrations,  and  we  have  pre- 
sented 12  in  a  period  of  one  hour  and  a  quarter;  40  would  take 
six  hours.     We  did  not  present  the  others. 


130  TRANSACTIONS  I.  E.  S. — PART  II 

ILLUMINATION  AND  EYESTRAIN.* 


ELLICE   M.  ALGER,   M.  D. 


Synopsis:  In  this  paper  the  following  topics  are  discussed  in  the 
order  given:  Nature  of  eyestrain  and  the  part  played  by  illumination — 
Daylight  the  ideal  illuminant — The  composition  of  artificial  lights  and 
the  effect  of  the  short  waves  in  producing  discomfort  and  disease  in  the 
eyes — The  intensity  of  light  and  its  lack  of  diffusion  as  factors  in  ocular 
fatigue  and  inefficiency — The  relation  between  poor  lighting  and  industrial 
accidents — The  arrangement  and  position  of  lights  and  their  influence  on 
efficiency — Experts  agreed  on  the  general  principles  involved  in  good 
lighting,  but  not  on  the  details — Education  and  discussion  must  precede 
any  drastic  legal  regulation. 

Eyestrain  is  the  common  expression  for  that  rather  compre- 
hensive group  of  symptoms  which  result  from  abnormal  ocular 
fatigue.  It  results  from  compelling  eyes  to  do  work  which  is 
beyond  their  physiological  capacity.  Things  close  at  hand  are 
seen  by  a  muscular  effort  of  focussing  which,  when  long  con- 
tinued, produces  a  normal  fatigue  and  requires  a  definite  period 
of  recuperation.  If  the  eyes  tire  sooner  than  they  should  because 
of  some  intrinsic  weakness  of  ciliary  muscles,  or  because  of  a 
handicap  imposed  by  astigmatism  or  increasing  years,  the  fatigue 
is  apt  to  manifest  itself  not  only  by  defective  vision  but  by  pain, 
and  we  speak  of  the  condition  as  an  accomodative  asthenopia. 
The  eye  likewise  sees  things  through  the  effect  of  light  falling  on 
a  sensitive  retina.  If  this  light  be  over  bright,  or  if  the  retina 
by  reason  of  over-exposure  or  disease  is  hypersensitive,  the 
result  is  the  disturbance  of  vision  and  pain  which  we  call  retinal 
asthenopia.  The  results  of  eyestrain  are  manifold  and  affect  no 
two  people  exactly  alike.  They  include  pains  in  the  eyes  and  many 
functionl  defects  of  vision  and,  quite  possibly,  often  result  in 
organic  eye  disease  as  well.  They  cause  80  per  cent,  of  the 
chronic  headaches.  They  often  result  in  functional  disturbances 
of  other  organs  and  in  conditions  of  general  nervous  exhaustion 
and  irritability.  While  most  of  the  symptoms  that  come  from 
eyestrain  are  of  the  accomodative  sort,  these  are  all  capable  of 
aggravation  by  improper  lighting  and  there  are  so  many  that  are 

*  A  paper  read  before  a  meeting  of  the  New  York  section  of  the  Illuminating  Engi- 
neering Society,  March  13,  1913. 


AivGER:     illumination  and  eyestrain  131 

caused  directly  in  this  way  that  I  shall  invite  your  attention  for 
a  time  to  the  relation  between  eyestrain  and  illumination. 

In  studying  this  relation  it  must  be  remembered  that  there  are 
few  exact  standards  of  ocular  capacity.  The  average  individual 
can  see  objects  of  a  definite  size  at  a  definite  distance  and  this 
average  is  taken  as  a  standard,  but  there  are  many  who  fall  below 
this  standard  without  obvious  cause,  and  many  who  are  far  above 
the  average.  The  variations  in  muscular  endurance  are  still 
wider  and  one  man  can  work  hour  after  hour  at  tasks  which 
fatigue  another  in  a  very  short  time.  The  sensitiveness  to  light 
likewise  varies  widely  in  different  individuals,  both  ability  to  see 
distinctly  by  faulty  light  and  ability  to  work  without  exhaustion 
in  strong  light. 

Light  is  the  reaction  excited  in  the  retina  by  the  impact  of 
certain  vibrations  or  waves  in  the  ether,  which  cause  different 
sensations  according  as  they  are  longer  or  shorter.  The  long 
ones  give  the  sensation  of  red  light,  while,  as  they  get  shorter 
and  shorter,  one  may  see  in  succession  all  the  colors  of  the  visible 
spectrum.  The  mixture  of  all  these  wave-lengths  together  pro- 
duces the  sensation  of  white  light.  But  the  visible  spectrum  does 
not  include  all  the  waves  by  any  means.  There  are  longer  waves 
than  the  red  which  cannot  be  seen  but  can  be  felt  as  heat,  and 
shorter  ones  than  the  violet  which  have  a  very  active  chemical 
effect.  It  must  be  remembered  too  that  both  these  qualities 
exist  in  the  visible  spectrum,  the  heating  effects  predominating 
at  the  red  end,  while  the  violet  end  approaches  the  ultra-violet  in 
its  chemical  activity. 

This  enables  one  to  explain  some  of  the  untoward  effects  of 
daylight  on  the  eyes,  even  though  daylight  affords  the  best  illumi- 
nation for  ordinary  purposes. 

Many  of  the  effects  of  sunlight  which  were  once  attributed  to 
heat  are  now  known  to  be  due  to  chemical  activity.  For  instance, 
in  snow  and  desert  blindness  the  light  is  broken  up  by  reflection 
from  the  crystalline  snow  or  sand,  and  the  actinic  waves  produce 
intense  inflammation  of  the  conjunctiva  which,  if  long  continued, 
results  in  total  disability.  Even  in  temperate  climes  one  suffers 
more  or  less  from  glare  and  burn  from  direct  or  reflected  sun- 
light, and  by  common  consent  a  good  north  light  is  taken  as  the 
standard  of  ideal  illumination,  being  the  steadiest,  the  pleasantest 


I32  TRANSACTIONS  I.  E.  S. — PART  II 

to  the  eyes,  the  best  diffused,  causing  the  fewest  shadows  and 
affecting  color  values  least. 

An  artificial  light  can  be  broken  up  into  its  component  parts 
and  its  spectrum  compared  with  that  of  daylight,  and  its  illuminat- 
ing power  can  be  measured  by  aid  of  various  photometers ;  but 
so  far  there  is  no  artificial  light  which  is  just  like  daylight,  though 
we  are  said  to  be  getting  nearer  and  nearer  to  it. 

It  has  been  shown  by  experiment  that  the  light  which  gives  the 
maximum  of  illumination  with  the  minimum  of  irritation  of  the 
eye  is  composed  of  the  yellowish  rays  from  the  middle  of  the 
spectrum.  For  this  reason  the  old  fashioned  candle  and  kerosene 
lights  have  never  gone  entirely  out  of  fashion.  But  most  of  the 
more  recent  artificial  lights,  whether  gas  or  electric,  contain  a  much 
higher  proportion  of  the  short  violet  or  actinic  rays  and  some  of 
them  contain  many  of  the  ultra-violet  rays  as  well.  When  un- 
shaded their  chemical  activity  is  so  great  that  they  can  be  used 
for  various  therapeutic  purposes.  They  are  capable  of  tanning 
the  skin,  and  of  causing  symptoms  like  those  of  a  modified  snow 
blindness.  Prolonged  exposure  to  the  electric  arc  light  some- 
times produces  an  intense  conjunctivitis  with  contraction  of  the 
pupils  and  erosions  of  the  cornea,  which  fortunately  generally 
yield  readily  to  treatment.  Nearly  everybody  has  experienced  the 
discomfort  and  premature  fatigue  that  comes  from  reading  by  un- 
shaded incandescent  lights.  Even  if  they  do  not  actually  produce 
inflammatory  changes  themselves,  they  certainly  render  those  al- 
ready present  decidedly  less  tolerable. 

It  is  quite  possible,  however,  that  the  delayed  actinic  effects  of 
light  whether  natural  or  artificial  are  much  more  serious.  The 
ultra-violet  rays  are  arrested  by  ordinary  glass,  and  in  the  eye  by 
the  tissues  of  the  cornea  and  lens  so  that  the  deeper  structures  of 
the  eye  escape  harm,  but  there  is  strong  reason  to  suspect  that 
their  constant  absorption  by  the  lens  may  be  one  of  the  causes  of 
cataract.  Experimenters  have  been  able  to  demonstrate  lenticu- 
lar changes  in  the  eyes  of  rabbits  exposed  to  such  lights;  and  it  is 
known  that  stokers  and  glass-blowers,  who  have  to  face  very 
brilliant  incandescent  light,,  have  a  tremendous  predisposition  to 
cataract.  This  so  called  bottle-makers'  cataract  begins  not  in  the 
anterior  part  of  the  lens,  which  would  be  expected  if  heat  were 


ALGER :      ILLUMINATION  AND  EYESTRAIN  133 

the  essential  factor,  but  in  the  posterior  portion  where  the  rays  of 
light  are  most  concentrated.  Other  suggested  observations  are 
that  in  the  ordinary  cataract  of  old  people  the  first  changes  gen- 
erally occur  in  the  lower  inner  quadrant  of  each  lens,  which  is  the 
part  least  shaded  by  the  brows  and  so  most  exposed  to  sunlight 
from  above,  and  that  when  cataracts  develop  in  people  who  have 
one  light  and  one  dark  eye,  it  invariably  appears  first  in  the  one 
unprotected  by  pigment  from  the  light. 

Even  if  the  ultra-violet  rays  do  not  reach  the  deeper  structures 
of  the  eye,  one  must  not  forget  that  the  shorter  waves  of  the  visible 
spectrum  have  decided  actinic  properties.  Many  people  have  had 
their  eyes  permanently  ruined  by  incautious  watching  of  an 
eclipse,  and  similar  damage  sometimes  follows  exposure  to  electric 
flashes  and  even  to  long  exposure  to  the  arc  light.  In  such  cases 
the  light  is  condensed  on  the  surface  of  the  retina  resulting  in 
local  inflammation  and  degeneration,  that  particular  spot  be- 
coming permanently  blind. 

Oculists  suspect,  though  they  cannot  prove,  that  less  intense  and 
longer  continued  light  irritation  may  be  a  factor  in  many  similar 
degenerative  changes  in  the  retina  and  chorioid,  and  advise  both 
for  prophylaxis  and  treatment  the  use  of  amber  glasses,  and 
shades  of  such  composition  as  to  soften  the  light  and  exclude  the 
actinic  end  of  the  spectrum.  To  people  who  are  at  all  sensitive 
to  light  they  are  a  great  comfort. 

Our  north  light  is  soft  and  even  and  well  diffused,  so  that  it 
causes  a  minimum  of  shadows.  Artificial  light  to  give  anything 
like  the  same  amount  of  illumination  must  be  much  more  con- 
centrated and  intense.  Now  the  human  eye  even  in  natural  light 
has  to  adapt  itself  to  so  many  variations  of  intensity  and  dimness 
that  it  has  developed  a  very  beautiful  mechanism  for  regulating 
the  amount  of  light  admitted  to  the  retina.  When  the  light  is 
dim  the  pupil  dilates,  and  when  it  is  bright  it  contracts  sharply.  A 
sudden  very  bright  light  causes  pain  not  because  the  retina  hurts 
but  because  of  this  sudden  extreme  muscular  contraction  of  the 
iris.  Constant  exposure  to  bright  light  necessitates  constant  muscu- 
lar contraction  and  engenders  in  many  people  premature  fatigue. 
Still  more  tiresome  and  painful  is  the  rapid  dilation  and  contrac- 
tion of  the  pupil  that  results  from  the  varying  intensity  of  a 


134  TRANSACTIONS  I.  E.  S. PART  II 

flickering  light.  Furthermore,  intense  and  long  continued  ex- 
posure to  bright  light  causes  retinal  exhaustion  and  the  retina  is 
capable  of  reacting  only  to  powerful  stimulation.  In  other  words, 
that  retina  becomes  for  the  time  being  blind  except  in  the  brightest 
of  lights.  Every  one  has  experienced  the  comparative  blindness 
caused  by  going  from  bright  sunlight  into  a  dimly  lighted  room. 
It  is  a  common  experience  to  have  workmen  insist  on  having  as 
intense  a  light  as  possible  because  they  have  temporarily  so 
blunted  their  retinal  sensitiveness  that  they  are  helpless  without 
it,  and  it  is  generally  the  hardest  kind  of  a  task  to  convince  them 
that  even  if  they  suffer  no  harm  from  the  glare  they  cannot 
possibly  work  as  long  without  fatigue.  In  another  set  of  people 
the  retina  instead  of  being  blunted  becomes  hyperesthetic  and 
finally  almost  incapable  of  bearing  any  exposure  to  light  at  all. 
This  condition  is  seen  at  its  worst  in  hysterics,  when  it  is  of  course 
not  a  result  of  over  lighting,  but  there  are  a  number  of  occupa- 
tions like  those  of  the  gilders  and  polishers  who  have  their  atten- 
tion fixed  for  long  periods  on  bright  surfaces,  in  which  retinal 
asthenopia  is  very  common. 

Furthermore,  daylight  has  a  vast  volume  and  is  dif- 
fused so  that  objects  get  light  from  all  sides,  and  shadows 
are  reduced  to  a  minimum.  Artificial  light  can  hardly 
be  expected  to  secure  thorough  diffusion  and  more  than  compara- 
tive freedom  from  shadows,  but  in  many  industries  almost  no 
attention  has  been  paid  to  this  point.  And  yet  it  is  very  im- 
portant, for  Calder  in  a  very  interesting  paper  has  shown  that 
the  retinal  anesthesia  and  deep  shadows  that  result  from  poor 
artificial  lighting  are  potent  factors  in  causing  industrial  accidents. 
The  records  of  some  8,000  manufacturing  plants  over  a  period 
of  three  years  showed  a  regular  minimum  of  accidents  during 
July  and  August  which  gradually  increased  to  a  maximum  in 
the  dark  winter  months.  The  influence  of  daylight  in  preventing 
accident  was  much  more  evident  in  occupations  which  require 
not  so  much  bright  light  as  diffused  light  without  shadows  over 
large  areas  as  in  the  building  trades  for  instance.  Indoor  work- 
ers as  often  suffer  accident  from  too  much  light  as  from  too 
little.  Exact  photometric  measurements  often  show  that  the  light 
of  ordinary  incandescent  lamps  concentrated  at  the  cutting- point 


ALGER :      ILLUMINATION  AND  EYESTRAIN  135, 

of  a  tool  or  a  work-bench  is  often  several  times  the  intensity  of 
daylight.  But  the  eye  adopts  itself  to  this  intensity  and  when 
the  workman  turns  from  his  over-lighted  work,  perhaps  in  a 
room  full  of  moving  machinery,  he  is  practically  blind.  What  is 
needed  from  the  point  not  only  of  safety  but  of  health  and  com- 
fort is  much  less  intensity  and  much  better  diffusion  of  light. 
This  applies  to  all  walks  of  life.  We  have  all  become  accustomed 
to  using  far  more  intense  light  than  we  need. 

One  can  measure  the  amount  of  illumination,  by  photometers 
which  are  much  more  accurate  and  dependable  than  the  human 
eye;  but,  after  all  is  said  and  done,  the  eye  is  one  of  the  best  of 
photometers  if  one  is  careful  not  to  injure  it  in  the  process,  since 
it  is  upon  its  adaptability  to  that  eye  that  all  artificial  lights  must 
stand  or  fall.  One  should  begin  with  a  low  illumination  and 
gradually  increase  it  till  a  point  is  reached  when  further  increase 
ceases  to  improve  the  details  of  the  work  in  hand.  Beyond  this 
all  additional  light  is  both  unnecessary  and  physiological  extrava- 
gance. 

Abnormal  fatigue  is  admittedly  one  of  the  greatest  predis- 
posing causes  to  most  diseases  be  they  physical  or  mental,  and 
though  the  part  played  by  bad  lighting  is  perhaps  not  clear  cut 
it  is  beyond  doubt.  In  most  factories,  schools  and  offices  the  eyes 
must  be  used  constantly  for  work  of  a  character  they  were  never 
intended  for.  The  result  even  in  normal  eyes  is  a  muscular  and 
nervous  fatigue  which  is  measurably  increased  by  both  over  or 
under-lighting.  The  first  engenders  fatigue  from  retinal  exhaus- 
tion and  pupillary  spasm,  while  the  second  results  in  the  strain 
that  follows  sharp  focussing  and  constant  attention.  In  the 
majority  of  individuals  whose  eyes  are  handicapped  by  astig- 
matism or  other  refractive  errors,  the  strain  is  still  greater.  I 
shall  not  take  your  time  with  the  long  list  of  conditions  of  health 
which  have  been  attributed  to  eyestrain.  Some  of  these  are 
beyond  dispute,  others  are  still  in  question. 

The  over-lighting  which  is  so  common  to-day  may  conceivably 
have  other  effects.  Woodruff  has  shown  that  in  the  tropics 
blondes  who  are  unprotected  by  skin  pigment  are  over-stimulated 
by  the  bright  light  and  finally  develop  a  characteristic  nervous 
exhaustion.  It  is  quite  possible  that  eyestrain  and  the  constant 
exposure  to  intense  light  of  short   wave-length  may  be  predis- 


136  TRANSACTIONS  I.  E.  S. — PART  II 

posing  factors  to  the  neurasthenia  from  which  our  garment 
makers  admittedly  suffer. 

The  arrangement  of  lights  as  well  as  their  composition  and 
intensity  is  of  importance.  It  is  well  known  how  uncomfortable 
it  is,  and  how  much  it  interferes  with  clear  vision  to  have  a  bright 
light  shining  directly  into  the  eyes,  and  lights  which  enter  the 
eye  from  below  are  much  more  annoying  than  those  from  above. 
And  yet  how  often  are  machines  so  placed  that  the  operator  has 
to  face  a  window  or  a  light.  The  same  difficulty  occurs  in  trades 
like  those  of  the  gilders  and  polishers  who  have  bright  lights 
reflected  into  the  eyes  from  their  work,  and  in  schools  where 
the  smooth  shining  pages  of  the  books  answer  the  same  purpose. 
So  far  as  possible  light  should  fall  from  above,  behind,  and  to 
one  side.  The  light  should  be  sufficient  for  the  work  in  hand, 
should  throw  no  shadows  on  the  work,  and  should  be  reflected, 
not  into  the  eyes  of  the  workers  but  to  one  side.  When  it 
comes  to  the  arrangement  of  light  for  many  workers  in  a  factory 
or  school  the  problem  is  very  much  more  difficult  and  presents 
many  technical  details,  which  must  be  left  in  the  hands  of  the 
illuminating  engineer. 

Even  when  estimated  by  its  actual  cost  in  dollars  and  cents,  bad 
lighting  is  often  more  expensive  than  good,  but  from  the  stand- 
point of  efficiency  there  is  no  comparison.  Bad  lighting  un- 
doubtedly causes  unnecessary  strain  of  the  eyes  and  consequent 
premature  muscular  fatigue;  it  compels  closer  and  more  con- 
stant attention  to  the  details  of  work,  so  that  tasks  which 
should  be  done  almost  automatically  and  without  mental 
effort  are  done  consciously.  Under  such  circumstances 
the  output  of  each  individual  is  manifestly  less  than  it  should  be, 
there  is  a  larger  percentage  of  mistakes  and  material  spoiled,  and 
the  number  of  accidents,  large  and  small,  is  vastly  increased. 
Even  under  the  best  of  conditions,  the  extreme  subdivision  of 
factory  work  with  its  consequent  monotony,  largely  destroys  the 
pleasure  of  work,  but  bad  eyes  and  poor  lighting  and  long  hours 
are  important  factors  in  the  industrial  discontent  of  the  day. 

It  goes  without  saying  that  any  system  of  scientific  shop  man- 
agement worthy  of  the  name  implies  a  good  lighting  system  as 
one  of  the  first  requisites,  but  as  yet  opinions  vary  widely  as  to 
just  what  this  means.     It  is  possible  to  regulate  the  color  and 


ALGER :       ILLUMINATION  AND  EYESTRAIN  137 

composition  of  the  light  that  enters  the  eyes  by  the  interposition 
of  screens  or  shades  which  shall  absorb  the  rays  one  does  not 
wish  to  use,  or  by  having  it  reflected  from  suitable  colored  sur- 
faces. The  volume  and  intensity  of  light  can  be  regulated  by 
increasing  or  diminishing  the  number  of  units,  and  by  diffusing 
it  with  frosted  shades,  or  by  reflecting  it  from  rough  surfaces. 
But  while  the  experts  are  agreed  on  the  principles  involved  they 
do  not  agree  entirely  on  the  details.  The  human  eye  is  flexible 
enough  to  adapt  itself  to  very  wide  variations  in  illumination  but 
there  must  be  comparatively  narrow  limits  within  which  the 
greatest  efficiency  may  be  reached.  Quite  possibly  different 
industries  may  require  entirely  different  types  of  illumination 
and  while  these  may  be  worked  out  in  detail  in  the  laboratory 
they  must  all  be  subjected  to  the  final  test  in  the  shop  or  school. 

Illumination  as  a  science  is  yet  in  its  infancy.  Even  in  great 
public  buildings,  libraries,  and  theatres  it  is  treated  not  as  an 
essential  but  simply  as  an  aid  to  the  proper  display  of  the  genius 
of  the  architect  or  the  taste  of  the  decorator.  And  if  such  build- 
ings are  badly  done  one  can  hardly  expect  as  yet  that  any  great 
attention  will  be  paid  to  the  proper  lighting  of  the  ordinary 
factory  or  house. 

Every  one  admits  to-day  that  the  state  must  control  factory 
conditions  so  far  as  they  effect  the  health  and  well  being  of  em- 
ployees and  many  attempts  are  being  made  both  here  and  abroad 
to  deal  with  the  subject  of  illumination  by  law. 

It  is  an  extremely  difficult  subject  to  handle  in  this  way;  even 
the  experts  are  not  agreed  on  many  important  points.  What 
would  be  good  lighting  in  one  industry  might  be  the  worst  possi- 
ble in  another.  To  make  drastic  regulations  in  the  present  state 
of  the  art  would  often  involve  manufacturers  in  great  expense  in 
changing  their  light  equipment  without  any  guarantee  that  it 
would  be  permanently  satisfactory. 

After  all,  good  lighting  is  essential  to  the  efficiency  of  both 
employer  and  employees,  and  a  judicious  campaign  of  education 
will  make  them  both  appreciate  it.  Then  too,  there  are  numerous 
large  and  powerful  corporations  engaged  in  various  branches  of 
the  lighting  industry  and  there  is  perhaps  little  danger  that  the 
subject  will  be  allowed  to  be  forgotten  either  by  consumers  or 
legislators. 


I38  TRANSACTIONS  I.  E.  S. — PART  II 

DISCUSSION. 

Dr.  Percy  W.  Cobb  (communicated)  :  Dr.  Alger  has  said  that 
there  are  few  means  of  measuring  ocular  capacity,  to  which  I 
would  add  "which  are  of  practical  use  in  the  question  of  illumina- 
tion." 

To  the  ophthalmologist  the  test-types  are  familiar,  by  which 
the  size  of  the  smallest  type  that  can  be  read  furnishes  the  basis 
on  which  visual  acuity  is  measured.  Among  other  means  of 
estimating  the  capacity  of  the  eye,  one  is  by  its  perception  of 
small  differences  in  brightness  in  fairly  large  surfaces — to  be  dis- 
tinguished from  visual  acuity  estimation,  where  the  brightness- 
difference  between  object  and  background  is  large  and  the  size 
of  the  object  seen  is  minimal.  Similarly  the  color  sensitivity  of 
the  eye  can  be  measured  and  many  other  criteria  of  the  sensitive- 
ness of  vision  have  been  worked  upon,  but  never  made  practical 
for  the  purposes  which  concern  illuminating  engineers.  That 
branch  of  physiology  is  as  yet  in  its  infancy. 

The  major  disturbances  of  the  eyes,  such  as  ophthalmia  elec- 
trica,  snow  blindness,  occupation  cataract  and  the  like,  probably 
due  to  the  ultra-violet  radiation,  are  however  quite  frank  and  to 
the  ophthalmologist  easily  recognizable.  A  noteworthy  point  is 
that  these  disturbances  all  have  their  origin  in  conditions  far  re- 
moved from  those  of  customary  illumination.  On  the  other  hand, 
the  troublesome  minor  and  more  familiar  disturbances  which 
appear  in  the  use  of  ordinary  illuminants  (headache,  smarting  of 
the  eyeballs,  blurring  of  vision  and  so  on)  seem  to  take  place 
under  conditions  such  that  physical  considerations  practically  pre- 
clude the  ultra-violet  radiation  as  the  cause ;  for  we  know  that  the 
light  from  an  ordinary  tungsten  filament  is,  for  equal  visible  light, 
far  poorer  in  the  ultra  violet  light  than  daylight,  and  is  used  at 
illuminations  almost  incredibly  lower  than  those  that  obtain  in 
daylight  conditions.  With  a  good  daylight  illumination  in  a 
room,  turning  on  the  artificial  lights  makes  almost  no  impression 
on  the  eye,  indicating  how  relatively  low  the  artificial  illumina- 
tion really  is. 

It  is  another  matter  when  we  come  to  consider  such  a  thing  as 
a  light  source  in  the  visual  field.  Here  we  have  an  intense  focuss- 
ing of  energy  upon  a  minute  area  of  the  retina,  and  a  disturbance 


ILLUMINATION   AND   EYESTRAIN  139 

soon  results,  as  anyone  knows  who  has  glanced  at  a  naked  fila- 
ment for  a  second  or  two  and  noticed  the  after  images  of  it  which 
subsequently  disturb  his  vision.  Whether  the  ultra-violet  plays 
a  part,  this  is  an  open  question.  We  must  remember  that  the 
preponderance  of  the  visible  radiaton  is  extremely  large  in  cases 
which  the  illuminating  engineer  has  to  consider. 

Bearing  on  this  point  is  some  work  quite  recently  published  by 
Dr.  Ferree,  from  which  it  appears  that  the  eye  suffers  a  consider- 
able loss  in  power  from  work  under  a  system  of  artificial 
illumination,  of  which  a  number  of  the  sources  come  within  the 
field  of  vision.  The  loss  in  the  power  of  the  eye  appeared  not  in 
the  acuity  itself,  as  the  eye  was  able  to  see  equally  small  test- 
letters  after  the  period  of  work,  but  in  its  ability  to  maintain  that 
power;  the  identical  test-letters  viewed  after  the  work-period 
showed  a  greater  tendency  to  appear  blurred  at  times. 

Apart  from  high  intrinsic  brightness  of  objects  it  seems  much 
more  than  probable  that  widely  unequal  illuminations  in  different 
parts  of  the  room  can  induce  retinal  disturbance  and  eye-strain. 
The  writer  is  inclined  to  explain  the  minor  eye-disturbances  just 
mentioned  largely  on  this  basis.  The  eyes  working  in  an  over- 
done local  illumination,  on  a  bright  page  or  a  machine  with  its 
numerous  reflections,  are  turned  to  relatively  dark  places,  for 
which  they  are  wholly  unadapted,  and  back  again  to  the  intense 
light — for  which  by  that  time  they  are  again  unadapted.  It  is  not 
hard  to  imagine  how  such  conditions  can  induce  retinal  fatigue 
and  muscular  eye-strain.  A  more  tangible  loss  due  to  such  a 
system  is  the  economic  value  of  time  wasted  in  fumbling  for  a 
needed  article  in  a  dim  light  with  bright-adapted  eyes;  and  above 
all  the  danger  of  accident,  especially  in  manufacturing  plants,  to 
persons  obliged  to  move  about  in  relatively  dim  surroundings  in 
which  their  eyes,  recently  under  high  illumination  are  more  or 
less  blind. 

For  such  reasons  legislation  as  to  illumination  conditions  in 
factories,  schools  and  other  public  and  semi-public  places  is  an 
important  part  of  the  modern  movement  for  the  conservation  of 
human  resources,  and  it  can  be  expected  that  such  legislation  will 
progress  whenever  experts  in  illumination  can  make  it  clear  ex- 
4 


140  TRANSACTIONS  I.  E.  S. — PART  II 

actly  what  is  the  most  efficient  lighting,  when  all  things  are  con- 
sidered. 

Dr.  Sinclair  Tousky  (communicated)  :  The  paper  is  re- 
markably complete  and  accurate. 

Eye-strain  is  much  more  commonly  a  result  of  imperfection 
in  the  eye,  than  in  the  light.  The  muscles  which  regulate  the 
size  of  the  pupil  for  the  amount  of  light,  act  automatically 
and  practicably  without  fatigue  and  provide  for  a  wide 
latitude  in  the  intensity  of  illumination.  Where  eye-strain 
is  diagnosed  it  is  not  enough  to  correct  errors  in  illumination,  but 
the  eye  itself  should  be  tested  as  to  accommodation  or  focussing 
power,  as  to  astigmatism  or  difference  in  curvature  and  refractive 
power  along  different  meridians  of  the  eye,  and  as  to  the  ocular 
muscles  which  should  be  so  accurately  balanced  as  to  automati- 
cally direct  both  eyes  toward  the  same  object  without  fatigue. 

The  ultra-violet  rays  are  much  more  abundant  in  sunlight  upon 
the  mountain  top,  than  in  our  cities  where  they  have  been  filtered 
out  by  passage  through  additional  thousands  of  feet  of  air  and 
especially  the  dust  in  the  air.  Filtered  water  is  more  transparent 
to  the  ultra-violet  ray  than  the  air  we  breathe. 

Dr.  Alger  is  right  in  recommending  glasses  as  a  protection 
from  sun-light  upon  snow-fields  and  in  the  desert  and  for  electric 
arc  lights ;  and  it  should  be  added  that  the  worst  accidents  come 
from  the  blinding  flash  when  an  electric  current  is  short-circuited 
and  that  these  workers  also  should  be  protected  by  glasses. 

Glass  is  practically  opaque  to  the  ultra-violet  rays,  but  quartz 
or  "pebble"  lenses  are  transparent  to  them.  Colorless  glass  is  a 
complete  protection  from  the  invisible  ultra-violet  rays  but  of 
course  if  the  luminous  rays  are  too  powerful,  the  colored  glasses 
which  give  the  greatest  sense  of  rest,  amber  color  perhaps,  may 
be  used. 

I  use  glass  in  the  measurement  of  ultra-violet  content  in  light 
used  for  the  therapeutic  effect  of  those  rays.  The  light  from  a 
mercury-vapor  arc  in  a  quartz  tube  may  produce  a  certain  effect 
upon  a  photographic  film  in  a  small  fraction  of  the  time  required 
to  produce  the  same  effect  through  a  thin  piece  of  glass.  Such 
a  light  will  quickly  produce  a  severe  burn  upon  the  healthy  skin ; 
and  cures  the  terrible  disease  called  lupus,  but  is  manifestly  un- 


ILLUMINATION    AND   EYESTRAIN  I4I 

suited  for  ordinary  illumination.  A  mercury- vapor  light  in  a 
glass  tube  generates  an  equal  quantity  of  ultra-violet  rays  but 
they  cannot  escape  from  the  tube  and  have  no  harmful  effect. 
The  light  from  any  electric  arc  is  rich  in  ultra-violet  rays  and 
should  always  be  filtered  through  a  glass  shade  for  illuminating 
purposes. 

In  applying  my  test  for  ultra-violet  rays,  it  must  be  understood 
that  no  camera  lens  is  used,  the  film  or  sensitized  paper  is  ex- 
posed directly  to  the  light.  And  any  light  which  produces  a 
markedly  greater  photographic  effect  through  a  sheet  of  quartz 
crystal  than  through  a  sheet  of  glass  is  regarded  as  too  rich  in 
ultra-violet  rays  to  be  desirable  for  illuminating  purposes. 

Professor  Alger  quotes  Woodruff's  statement  that  it  is  not  the 
heat  but  the  light  in  the  tropics  which  injuriously  affects  the  white 
races.  This  theory  has  been  very  widely  tested  and  the  general 
conclusion  is  against  it.  So  that  instead  of  the  dark  underwear 
advised  by  Woodruff,  white  clothing  has  been  found  to  be  better. 

One  other  point  should  be  mentioned  and  that  is  the  eye-strain 
which  inevitably  results  from  a  flickering  light. 

Eye-strain  may  frequently  be  prevented  by  the  presence  of  a 
dimly  lighted  back-ground  upon  which  the  vision  may  rest  dur- 
ing moments  when  the  person  looks  away  from  his  work.  Con- 
stant staring  at  work  at  a  fixed  distance  and  with  either  too 
much  or  too  little  illumination,  will  cause  strain  in  the  best  of 
eyes. 

Dr.  C.  E.  FerrEE  (communicated)  :  Dr.  Alger's  paper  is  in 
brief  compass  an  excellent  and  interesting  symposium  of  the  sub- 
ject. The  present  writer  has,  however,  the  following  comments 
to  make. 

(1)  On  the  third  page  of  his  paper  Dr.  Alger  says:  "It  has 
been  shown  by  experiment  that  the  light  which  gives  the  maximum 
of  illumination  with  the  minimum  of  irritation  to  the  eye  is  com- 
posed of  the  yellowish  rays  of  the  middle  of  the  spectrum.  For 
this  reason  the  old-fashioned  candle  and  kerosene  lamp  have 
never  gone  entirely  out  of  fashion."  Although  he  willingly  leaves 
himself  open  to  correction,  the  present  writer  does  not  believe 
that  the  above  conclusion  can  be  justly  drawn  from  the  experi- 
mental evidence  in  existence  at  the  present  time.     There  are  two 


142  TRANSACTIONS  I.  E.  S. — PART  II 

points  to  Dr.  Alger's  comparison :  "maximum  illumination"  and 
"minimum  irritation."  Of  these  two  points,  so  far  as  the  writer 
knows,  definite  experimental  work  has  been  done  only  on  the 
former.1  By  "maximum  illumination"  the  writer  asumes  from 
the  relation  of  the  statement  to  the  general  subject  of  lighting 
that  Dr.  Alger  means  the  maximum  of  illumination  for  seeing 
detail  or  maximum  acuity  of  vision.2 

There  are  two  ways  in  which  acuity  of  vision  may  be  con- 
sidered in  relation  to  the  problem  of  lighting:  acuity  as  deter- 
mined by  the  momentary  judgment,  and  acuity  which  represents 
the  average  of  ability  to  see  detail  for  a  period  of  time.  Acuity 
as  determined  by  the  momentary  judgment  does  not  show  the 
progressive  loss  of  efficiency  resulting  from  a  period  of  work 
even  under  an  unfavorable  lighting  system,  because  under  the 
spur  of  the  will  the  muscles  of  the  eye,  though  they  may  have 
lost  enormously  in  efficiency,  may  be  whipped  up  to  their  normal 
power  long  enough  to  make  the  judgment  required  by  this  visual 
acuity  test.  Acuity,  then,  as  determined  by  the  momentary  judg- 
ment can  only  be  used  to  determine  the  general  level  or  scale  of 
efficiency  for  the  fresh  eye.  It  can  not  be  used  to  detect  loss  of 
efficiency.  In  the  problem  of  lighting,  however,  the  general  level 
or  scale  of  efficiency  as  determined  by  the  momentary  judgment 
of  acuity  is,  comparatively  speaking,  of  minor  importance.  What 
is  needed  is  a  type  of  illumination  that  gives  the  highest  average 
of  acuity  or  efficiency  in  seeing  for  a  period  of  work  and  at  the 
same  time  the  least  loss  of  efficiency.  So  far  as  its  relation  to 
the  quality  of  light  is  concerned,  tests  have  been  made  up  to  the 
present  time  only  of  acuity  as  determined  by  the  momentary  judg- 
ment.    With  regard  to  this  type  of  acuity  Dr.  Alger's  statement 

1  Although  no  definite  experimental  work  has  yet  been  done  on  the  effect  of  varying 
the  quality  of  light  on  its  tendency  to  produce  discomfort,  still  it  can  be  said  from  the 
results  of  our  own  work  that  when  intensity  and  distribution  are  equalized,  an  installa- 
tion of  clear  carbon  lamps,  which  gives  a  light  comparatively  rich  in  yellow  and  red,  shows 
a  greater  tendency  to  produce  discomfort  than  an  installation  of  clear  tungsten  lamps, 
the  light  from  which  contains  a  proportionately  greater  number  of  the  short  wave- 
lengths. 

-  Visual  acuity  as  usually  tested  involves  the  discrimination  both  of  visual  angle  and 
of  brightness  difference.  As  it  enters  into  seeing  in  ordinary  life,  it  may  involve  also  the 
discrimination  of  differences  in  color  quality. 


ILLUMINATION   AND   EYESTRAIN  143 

will  be  here  examined  in  the  light  of  the  work  done  by  three  men 
selected  as  typical :  Langley,3  Luckiesh,4  and  Rice.5 

Langley  made  his  determination  of  acuity  with  the  colors  of 
the  spectrum  equalized  in  energy.  Luckiesh  and  Rice,  on  the 
other  hand,  worked  with  colors  equalized  photometrically.  Some 
again  have  exercised  no  especial  intensity  control  at  all.  In  the 
writer's  opinion,  Langley's  conception  is  the  correct  one.  We 
want  to  know  for  equal  outputs  of  energy  what  color  gives  the 
greatest  acuity  of  vision.  To  equate  the  colors  photometrically 
is  to  equate  them  for  seeing,  which  in  a  measure  begs  the  ques- 
tion at  the  outset.0  Langley,  working  with  the  light  of  the 
spectrum,  showed  that  for  equal  amounts  of  energy  (radiometri- 
cally  determined)  the  maximal  acuity  of  vision  is  given  when 
light  in  the  region  of  the  green  and  blue-green  is  employed.  His 
results,  therefore,  give  no  support  to  the  belief  that  yellow  light 
possesses  an  advantage  over  lights  of  other  colors  for  clear  seeing. 

Luckiesh  was  not  concerned  primarily  with  making  a  com- 
parison of  the  different  colors  for  acuity,  although  such  a  com- 
parison may  be  made  from  his  results.  His  problem  was  to  show 
that  colors  taken  from  a  narrow  region  of  the  spectrum  give 
greater  acuity  than  colors  more  complex  as  to  wave-length,  and 
to  find  an  explanation  for  this  phenomenon.     An  examination, 

3  Langley:  Energy  and  Vision,  American  Journal  of  Science,  18S8,  XXXVI,  3rd  Ser., 
PP-  359-379- 

4  Luckiesh,  M.:  The  Influence  of  Spectral  Character  of  Light  on  the  Effectiveness  of 
Illumination,  Trans,  of  the  Illuminating  Engineering  Society,  1912,  VII,  p.  135-158. 

6  Rice,  D.  E.:  Visual  Acuity  with  Lights  of  Different  Colors  and  Intensities,  Arch,  of 
Psychol.,  1912,  No.  23,  p.  1-59. 

See  also  Uhtoff:  Archiv fur  Ophthalmologic  1S86,  XXXII,  (1),  p.  171.  A.  Konig:  Zeitsch. 
f.  Psychol,  u.  Physiol,  d.  Sinnesorgane,  1893,  IV,  p.  241;  and  Sitzungsbcr  d.  Berliner  Akad.  d. 
Wissensch.,  1897,  XIII,  p.  559;  Pfliiger:  Ann.  d.  Physik,  1902,  IX,  p.  185;  Oerum:  Skandi- 
navisches  Archiv  fur  Physiol.,  1904,  XVI;  Boltunow:  Zeitsch.  f.  Psychol,  u.  Physiol,  d. 
Sinnesorgane,  1907-8,  XLH,  (2),  p.  359;  Broca  and  Laporte:  Bulletin  de  la  Societe  Inter- 
nationale des  Electnciens,  Paris,  1908,  VIII;  2nd  Ser.,  No.  LXXVII;  Dow,  J.  S.:  London  III. 
Eng..  II,  p.  233;  Ashe,  S.  W. :  Electrical  World,  Feb.  25,  1909. 

6  That  is,  visual  acuity  as  ordinarily  tested  involves  a  discrimination  of  brightness 
difference  as  well  as  of  visual  angle.  The  discrimination  of  blightness  difference  sustains 
a  relation  both  to  the  degree  of  illumination  and  to  the  color  quality  of  the  light.  Up  to 
a  certain  point  an  increase  in  the  degree  of  intensity  of  illumination  increases  the  dis- 
crimination of  brightness  difference.  The  presence  of  dominant  color  in  the  light,  on  the 
other  hand,  interferes  with  this  discrimination  or  masks  the  difference.  If,  then,  the 
colored  lights  are  made  of  the  same  degree  of  luminosity  as  determined  by  the  photo- 
metric judgment,  there  remains  only  one  of  the  above  factors  to  make  them  differ  in  the 
degree  of  acuity  they  produce,  namely,  the  effect  of  color  quality  on  the  discrimination  of 
brightness  difference.  Therefore,  I  have  said  that  to  make  a  comparison  of  visual  acuity 
for  the  different  colors  with  lights  equated  photometrically  is  in  a  measure  to  beg  the 
question  at  the  outset. 


144  TRANSACTIONS  I.  E.  S. — PART  II 

of  his  curves  show  a  greater  acuity  for  yellow  than  for  any  of 
the  other  colors  of  the  spectrum.  But  a  comparison  can  not  be 
drawn  from  his  results  of  acuity  for  yellow  and  white  light;  it 
can  be  made  merely  for  yellow  and  the  other  colors  of  the  spec- 
trum. Moreover  it  is  far  from  safe  to  pass  from  the  results  of 
his  experiments  to  the  conclusion  that  yellow  light  gives  greater 
acuity  than  white  light;  and  still  less  safe  from  a  literal  inter- 
pretation of  these  results  to  conclude  that  white  light  with  yellow 
as  its  dominant  hue,  such  as  is  given  by  the  kerosene  flame,  gives 
greater  acuity  than  is  given  by  clear  white  light.  Yet  we  presume 
that  either  results  of  this  kind  or  results  of  photometric  observa- 
tions which  strictly  speaking  are  not  applicable  to  the  point  in 
question,  are  responsible  for  the  belief,  somewhat  generally  held, 
that  yellow  light  possesses  an  advantage  for  seeing  over  white  light. 
Theoretically  speaking,  this  belief  might  at  first  thought  seem 
to  have  considerable  justification,  for  if  white  is  made  up  of  all 
the  colors,  and  of  these  colors  yellow  gives  the  greatest  acuity,  then 
yellow  should  also,  it  might  seem,  give  greater  acuity  than  white 
light  of  equal  luminosity.  Too  much  would,  however,  be  taken  for 
granted  in  drawing  such  a  conclusion,  for  as  stated  in  the  pre- 
ceding foot  note,  visual  acuity,  as  ordinarily  tested,  involves  a  dis- 
crimination of  brightness  difference  as  well  as  a  discrimination 
of  visual  angle,  and  the  presence  of  a  dominant  color 
in  the  light  strongly  interferes  with  this  discrimination.  In 
Luckiesh's  experiments  this  color  factor  was  present  in  case  of 
each  of  the  comparisons  made.  Moreover,  it  was  probably 
weakest  in  case  of  yellow,  for  yellow  is  the  least  saturated  of  the 
colors  of  the  spectrum.  Hence,  yellow  in  this  regard  possesses 
an  advantage  over  all  the  other  colors  for  the  clear  seeing  of  de- 
tails executed  in  white  and  black,  as  is  required  in  the  visual 
acuity  test.  In  the  case  of  yellow  vs.  white  light,  however,  the 
advantage  is  reversed.  No  color  factor  is  present  to  reduce 
acuity  in  the  determinations  for  white  light.7     In  short,  there- 

7  The  difference  between-the  results  of  Langley  and  L,uckiesh  was  doubtless  due  to  the 
difference  in  their  method  of  equating  their  colors.  Since  Langley  equalized  his  colors  in 
energy,  he  had  differences  both  in  luminosity,  and  in  saturation  and  quality  of  color  to 
affect  his  discrimination  of  the  brightness  difference  between  his  test  object  and  its  back- 
ground for  the  different  colors;  while  Luekiesh,  since  his  colors  were  equated  in  lumin- 
osity, had  only  difference  in  color  quality  and  saturation  to  affect  this  discrimination. 
Possibly  resolving  power  of  the  lens  should  also  be  included  as  one  of  the  factors  influ- 
encing acuity  for  the  different  colors;  but  since  the  resolving  power  according  to  Ravleigh 
is  greatest  in  the  blue,  resolving  power  could  not  at  least  have  been  a  dominating  "factor 
in  I,uckiesh's  observations. 


ILLUMINATION    AND   EYESTRAIN  I45 

fore,  to  reason  without  further  experiment  from  results  of  the 
kind  obtained  by  Luckiesh,  to  what  would  happen  in  case  a  com- 
parison should  be  made  for  any  of  the  colors  and  white  light,  as 
apparently  has  often  been  done,  is  to  pre-suppose  the  assumption 
of  a  degree  of  simplicity  with  regard  to  the  eye  factors  that  does 
not  exist. 

Rice  did  not  work  with  the  light  of  the  spectrum.  He  used  the 
rougher  method  of  isolating  his  wave-lengths  afforded  by  color 
filters.  He  determined  the  acuity  for  light  approximating  white, 
(given  in  some  cases  by  a  carbon,  and  in  some  cases  by  a  Nernst 
filament),  and  for  the  colors  given  by  his  red,  green,  and  blue 
filters.  These  colors  were  made  photometrically  equal,  each  to 
each  and  to  the  white  light.  Ten  different  degrees  of  intensity 
were  employed.  The  acuities  for  these  lights  were  found  to  be 
in  order  from  greatest  to  least :  white,  red.  green,  and  blue. 
Yellow  did  not  enter  into  the  comparison.8 

Even,  then,  with  regard  to  acuity  of  vision  as  determined  by 
the  momentary  judgment,  it  has  not  been  established  by  experi- 
ment, so  far  as  the  writer  knows,  that  either  monochromatic  yel- 
low or  white  light  predominantly  yellow  possesses  an  advantage 
for  clear  seeing  over  clear  white  light ;  and  with  regard  to  a 
visual  acuity  determination  which  represents  an  average  for  a 
period  of  work,  no  experiments  have  as  yet  been  made  which 
can  be  regarded  as  determining  the  effect  of  quality  of  light  on 
clear  seeing.  In  his  own  experiments  the  writer  has  planned  to 
determine  in  order  the  effect  of  differences  in  distribution, 
intensity,  and  quality  of  light,  both  on  the  power  of  the  eye  to 
hold  its  efficiency  during  a  period  of  work,  and  to  maintain  its 
maximum  state  of  comfort.  The  effect  of  differences  in  dis- 
tribution and  intensity  is  now  being  worked  out.  The  results  for 
distribution  have  in  part  been  published,9  and  more  will  soon  be 

8  Oerum  (op.  cit.)  also  made  a  comparison  of  acuitv  for  white  light  and  the  colors. 
He  found  white  light  to  give  the  greatest  acuity.  Of  the  colors,  red,  green,  and  blue,  red 
gave  the  greatest:  green  next;  and  blue  the  least.  Boltunow  (op.  cit.),  however,  making 
a  similar  comparison,  found  green  to  give  the  greatest 'acuity  of  these  three  colors,  and 
red  the  least.  White  gave  a  greater  acuity  than  an3'  of  the  colors.  Ashe  (op.  cit.)  de- 
termined the  acuity  for  red,  green,  blue,  and  white  lights  of  equal  luminosities.  He 
found  the  greatest  acuity  for  white  lights.  Of  the  colors,  blue  gave  the  greatest  acuity, 
green  next,  and  red  the  least.  Dow  (op.  cit.)  found  that  light  in  the  region  of  the  blue- 
green  gives  the  greatest  acuity  for  near  objects,  and  light  in  the  region  of  the  red  for  dis- 
tant objects. 

9  C.  E.  Ferree.  Tests  for  the  Efficiency  of  the  Eve,  etc.  Trans.  I.  E.  S.  (Jan.,  1913), 
vol.  VIII.,  p.  40. 


I46  TRANSACTIONS  I.  E.  S. — PART  II 

published.  A  systematic  study  of  the  effect  of  differences  in 
quality  has,  however,  not  yet  been  undertaken.  The  writer  is, 
therefore,  not  at  this  time  in  a  position  definitely  to  commit  him- 
self on  this  point.  He  can  say,  however,  from  the  results  of  the 
work  already  done,  that  with  distribution  and  intensity  equated, 
an  installation  of  clear  carbon  lamps,  which  gives  a  light  rela- 
tively rich  in  red  and  yellow,  causes  the  eye  to  fall  off  more  in 
efficiency  as  the  result  of  a  period  of  work  than  an  installation  of 
clear  tungsten  lamps,  the  light  from  which  is  whiter  and  contains 
proportionately  more  of  the  shorter  wave-lengths.  In  short,  it  is 
the  writer's  contention  that  the  question  whether  or  not  white  or 
colored  light  is  better  for  the  eye  can  not  be  answered  until  definite 
tests  are  made  of  this  point  alone  under  conditions  in  which  all 
other  factors  are  rendered  constant.  The  merits  of  the  kerosene 
flame,  for  example,  as  compared  with  other  sources  of  illumina- 
tion, must  be  tested  under  a  system  of  installation  that  gives  the 
same  intensity  at  the  source  and,  as  nearly  as  possible,  the  same 
distribution  in  the  field  of  vision  as  is  given  by  other  illuminants. 
This  has  not  been  done  at  all.  Our  judgment  of  the  comparative 
merits  of  the  color  quality  of  the  light  given  by  it  are  based  on 
the  roughest  kinds  of  impression  obtained  under  conditions  of 
installation  in  which  there  has  been  no  attempt  at  control  of  the 
other  factors  that  influence  the  effect  of  light  on  the  eye. 

(2)  On  the  sixth  page  of  his  paper  Dr.  Alger  says:  "One 
should  begin  with  a  low  illumination  and  gradually  increase  it 
until  a  point  is  reached  where  further  increase  ceases  to  improve 
the  details  of  the  work  in  hand."  That  is,  visual  acuity  is  here 
made  the  test  of  the  amount  of  light  that  should  be  employed. 
But,  as  is  stated  above,  visual  acuity,  that  is  visual  acuity  as 
determined  by  the  momentary  judgment,  is  not  the  only  or  even 
the  most  important  factor  that  has  to  be  taken  into  account  in 
lighting.  The  element  of  time  must  be  introduced  into  the  test. 
That  is,  an  intensity  of  illumination  must  be  chosen  at  which  the 
eye  holds  its  maximum  acuity  or  efficiency  for  a  period  of  work. 
This  is  by  no  means  in  every  case  the  same  degree  of  illumination 
that  gives  the  maximum  of  acuity  as  determined  by  the 
momentary  judgment.  For  example,  our  tests  for  loss  of 
efficiencv  for  different  intensities  of  illumination  with  a  given 


ILLUMINATION   AND   EYESTRAIN  147 

type  of  installation  do  not  show  that  the  degree  of  illumination 
that  gives  maximum  acuity  for  the  momentary  judgment  gives 
also  for  every  kind  of  installation  the  least  loss  of  acuity  or  the 
maximum  average  of  efficiency  for  a  period  of  3-4  hours  of  work. 
Moreover,  the  comfort  of  the  eye  must  also  be  taken  into  account. 
Here  also  results  are  wanting  to  show  that  the  degree  of  illumi- 
nation that  gives  maximum  acuity  as  determined  by  the 
momentary  judgment  gives  also  maximum  comfort.  In  short, 
the  degree  of  correlation  between  visual  acuity,  loss  of  efficiency 
and  the  tendency  to  produce  discomfort  can  not  be  taken  for 
granted.  These  factors  constitute  three  separably  determinable 
moments,  no  one  of  which  should  be  neglected  in  installing  a 
lighting  system. 

(3)  Dr.  Alger  says  on  the  eight  page:  "The  human  eye  is 
flexible  enough  to  adapt  itself  to  very  wide  variations  of  illumina- 
tions, but  there  must  be  comparatively  narrow  limits  within  which 
the  greatest  efficiency  may  be  reached."  Our  work  on  the  effect 
of  varying  the  intensity  of  illumination  on  the  eye's  loss  of 
efficiency  shows  that  in  general  this  statement  is  true.  The  range 
of  favorable  intensity  varies  widely,  however,  with  the  type  of 
installation.  It  is,  for  example,  much  narrower  for  the  direct 
than  for  the  indirect  lighting  system. 

(4)  On  the  seventh  page  Dr.  Alger  says  that  "lights  which 
enter  the  eye  from  below  are  much  more  annoying  than  those 
from  above."  This  statement  is  also  borne  out  by  the  results  of 
our  experiments.  In  making  a  preliminary  study  of  the  causes 
of  discomfort,  a  light  of  constant  intensity  was  thrown  on  the 
retina  at  different  points  in  its  several  meridians,  and  the  time 
limen  of  discomfort  was  determined.  This  limen  was  found  to 
be  lower  for  the  upper  than  for  the  lower  half  of  the  retina;  and 
for  the  nasal  than  for  the  temporal  half. 

(5)  On  the  fourth  page  Dr.  Alger  says:  "A  sudden  very 
bright  light  causes  pain  not  because  the  retina  hurts,  but  because 
of  a  sudden  extreme  muscular  contraction  of  the  iris.  Constant 
exposure  to  bright  light  necessitates  constant  muscular  contrac- 
tion and  engenders  in  many  people  premature  fatigue." 
Magendie's  experiments  in  1824  showed  that  the  retina  and  optic 
nerve  are  insensitive  to  pain  from  mechanical  stimulation.  These 


I48  TRANSACTIONS  I.  E.  S. — PART  II 

and  similar  experiments  have  led  to  the  belief  that  the  discomfort 
experienced  on  exposing  the  eye  to  a  degree  of  illumination  to 
which  it  is  not  accustomed  is  muscular.  That  it  can  not  be 
wholly  or  even  essentially  muscular  is  shown  by  the  fact  that  it 
is  gotten  in  cases  where  the  ciliary  and  iris  muscles  have  been 
paralysed  by  atropin;  also  in  cases  where  the  lens  has  so  long 
been  removed  that  muscular  atrophy  must  have  taken  place.  In 
short,  there  is  no  doubt  in  the  present  writer's  mind  that  the  dis- 
comfort experienced  as  the  result  of  work  under  unfavorable 
conditions  of  lighting  is  not  by  any  means  all  muscular.  The 
"sandiness"  passing  over  into  a  stinging  stabbing  pain  which 
comes  early  in  the  experience  of  discomfort  seems  to  be  con- 
junctival. Just  what  other  reactions  come  as  the  result  of  ex- 
posing the  retina  to  a  degree  or  kind  of  illumination  to  which  it  is 
not  accustomed  is  for  future  work  to  determine.  That  they 
can  not  all  be  muscular  is  plainly  obvious. 


hunter:     home  lighting  149 

HOME  LIGHTING.* 


BY    GEORGE    LELAND    HUNTER. 


Synopsis:  The  following  paper  presents  some  of  the  elementary  con- 
siderations of  residence  or  home  lighting,  particularly  from  the  viewpoint  of 
the  interior  decorator.  The  author  regards  home  lighting  as  a  decorative 
problem.  He  discusses  in  a  general  way  the  influence  of  colored  surfaces 
and  contrast  upon  the  appearance  and  the  decorative  scheme  of  interiors.  He 
suggests  the  co-operation  of  the  architect,  illuminating  engineer  and 
decorator  to  insure  success  and  progress  in  lighting  problems  of  this  kind. 

In  the  houses  and  apartments  of  New  York  and  other  large 
cities  there  are  thousands  of  rooms  gloomy  and  cheerless  because 
they  are  not  properly  lighted,  either  by  day  or  by  night.  Because 
of  such  rooms  hundreds  of  flats  and  apartments  are  unrented 
which  if  properly  illuminated  could  be  quickly  leased  to  respon- 
sible and  permanent  tenants. 

The  first  impulse  is  to  rush  to  the  lighting  fixture  man  or  to 
the  illuminating  engineer,  for  assistance  in  solving  the  problem. 
From  the  name,  people  naturally  assume  that  an  illuminating 
engineer  is  one  capable  of  producing  ideal  conditions  of  illumi- 
nation in  any  kind  of  an  interior.  As  a  matter  of  fact  the  illumi- 
nating engineer  has  not  always  a  remedy  at  hand  to  correct  the 
bad  lighting  of  dark  rooms  and  apartments.  Nor  is  the  lighting 
fixture  man  any  better  off.  They  can  talk  about  fixtures,  candle- 
power,  frosted  bulbs  and  diffusing  shades  and  the  distribution 
of  outlets,  etc.,  but  when  assigned  a  problem  in  lighting  a 
dark  room  agreeably  and  effectively  they  often  fail.  This  is  not 
their  fault.  The  specialization  of  industry  prevents  most  light- 
ing fixture  men  from  knowing  anything  about  interior  decoration 
and  furnishings. 

The  problem  of  illuminating  the  interiors  of  houses  and  apart- 
ments is  not  only  a  fixture,  but  a  decorative  problem.  Success 
in  the  future  lies  in  the  co-operation  of  architect,  decorator  and 
lighting  fixture  man.  Of  course  when  the  architect  really  under- 
stands the  art  of  interior  decoration,  as  well  as  the  planning  and 
construction  of  buildings,  his  is  the  master  mind  that  can  best 
guide  the  lighting  man  and  the  decorator. 

*  A  paper  read  before  a  meeting  of  the  New  York  section  of  the  Illuminating  Engi- 
neering Society,  December  12,  1912. 


150  transactions  i.  e.  s. — part  ii 

Illuminating  a  room  means  making  clearly  visible  the  form 
and  color  and  texture  of  the  objects  in  it.  If  the  walls,  furniture 
and  other  objects  in  a  room  are  properly  illuminated,  then  the 
room  is  illuminated.  If  they  are  not  properly  illuminated,  then 
the  room  is  not  properly  illuminated. 

Of  course  there  is  opportunity  for  all  kinds  of  contrast  in 
lighting  a  room.  Often  the  woodwork — the  architectural  back- 
ground, the  frame-work  of  the  room — can  be  accentuated  by 
being  kept  a  little  heavier  in  color  and  consequently  darker  than 
the  flat  surfaces  of  the  walls.  Often  too  the  furniture  is  very 
properly  accentuated  in  the  same  way.  But  if  all  the  objects 
in  a  room,  as  well  as  the  walls  and  ceiling  and  floor,  are  dark 
and  somber,  it  is  then  impossible  to  illuminate  the  room  at  all. 

The  only  thing  that  can  be  done  is  to  distribute  around  the 
room  a  large  number  of  light  units  having  comparatively  large 
shades  of  diffusing  glass  that  will  give  large  surfaces  of  bright- 
ness— but  surfaces  that  are  not  too  bright.  With  this  wide  dis- 
tribution of  light  sources,  wherever  the  eye  turns  it  meets  bright- 
ness, so  that  the  effect  is  one  of  brightness.  In  illumination  that 
is  everything. 

In  measuring  illumination,  the  photometer  is  practically  use- 
less. The  human  eye  is  the  only  capable  judge.  Whether  a 
room  is  photometrically  bright  is  of  not  the  slightest  importance. 
The  only  thing  of  importance  is  "Does  the  room  look  bright?" 
If  it  looks  bright  it  is  bright. 

It  is  a  well  known  fact  that  light  surfaces  reflect  much  light 
and  that  dark  surfaces  reflect  little  light.  This  means  that  the 
dark  surfaces  absorb  the  light,  while  the  light  surfaces  reflect  it. 
This  applies,  of  course,  not  only  to  the  walls  of  a  room  but  also 
to  the  windows  of  a  room.  In  illumination  this  fact  is  of  vital 
importance.  If  the  walls  of  a  room  are  light,  they  not  only 
look  bright  but  they  also  reflect  light  to  other  walls,  which 
re-reflect  the  light.  And  at  each  reflection  but  a  small  part, 
comparatively,  of  the  light  is  lost. 

I  have  noticed  that  many  people  in  the  evening  leave  the 
window  shades  up.  Apparently  it  does  not  occur  to  them  that 
every  open  window  is  a  transparent  hole  through  which  the  light 
leaks.    And  light  costs  money.    Dark  window  shades  also  absorb 


hunter:     home  lighting  151 

a  great  deal  of  light.  Light  shades  not  only  reflect  light  in  the 
room  but  by  their  brightness  help  give  an  appearance  of  bright- 
ness and  cheerfulness  to  the  whole  interior. 

While  it  is  true  that  light  surfaces  always  tend  to  recede  from 
the  eye  and  dark  surfaces  to  advance  toward  the  eye,  it  is  also 
true  that  small  spots  or  objects  tend  to  stand  out  against  a  con- 
trasting background.  But  bright  objects  against  a  dark  back- 
ground tend  to  stand  out  less  than  dark  objects  against  a  light 
background.  These  facts  are  of  prime  importance  in  the  decora- 
tive and  economical  lighting  of  a  room.  The  only  way  to  light 
residences  economically  is  to  light  them  decoratively.  If  they 
are  not  lighted  decoratively  they  are  not  lighted  at  all.  It  is 
lighting  that  makes  a  room  agreeable  and  comfortable  to  live  in. 

The  shape  and  size  of  rooms  are  also  of  vital  importance  in 
illuminating  them.  I  find  that  a  great  many  lighting  fixture  men 
do  not  seem  to  have  any  definite  relation  between  the  shape  and 
size  of  rooms  and  the  amount  of  light  they  provide.  I  will  admit 
that  the  amount  of  light  necessary  varies  very  greatly  according 
to  the  way  the  rooms  are  finished  and  furnished.  A  room  fin- 
ished and  furnished  in  dark  colors  with  surfaces  of  rough  texture 
might  require  ten  times  as  much  light  as  a  room  with  light  smooth 
surfaces,  and  not  be  illuminated  at  that. 

In  my  own  practise,  in  rooms  9  feet  high,  I  allow  a  16  candle- 
power  lamp  for  each  fifty  feet  of  floor  space.  In  higher  rooms 
I  allow  10  per  cent,  additional  light  for  each  increase  of  one 
foot  in  height.  This  will  be  found  too  much  light  for  light 
rooms,  but  too  little  for  dark  room. 

There  are  some  rooms  in  a  house  with  ceilings  too  high  for 
their  size — such  as  narrow  halls,  bath  rooms,  etc.  In  these  rooms 
ceiling  reflection  should  not  be  utilized.  The  moment  a  ceiling 
is  brightly  lighted  it  appears  to  rise  from  one  to  two  feet.  Con- 
sequently in  bath  rooms  and  narrow  halls  the  light  should  be 
kept  off  the  ceiling,  and  placed  upon  the  side  walls,  making 
the  side  walls  appear  to  recede  while  the  dark  ceiling  appears 
lower. 

Of  course,  if  a  room  is  very  wide  and  low,  it  is  desirable  to 
use  ceiling  reflection.  By  making  the  ceiling  bright,  its  height 
is  apparently  increased ;  at  the  same  time  it  may  be  used  as  a 
secondary  distributor  of  the  light. 


152  TRANSACTIONS  I.  E.  S. — PART  II 

Most  of  the  talk  about  indirect  illumination  is  nonsense.  I 
would  absolutely  bar  from  use  for  ceiling  fixtures  anything  that 
does  not  allow  a  fair  proportion  of  the  light  to  come  down 
through  the  fixture.  The  best  fixtures  for  what  is  called  semi- 
indirect  lighting  are  alabaster  bowls  and  bowls  of  a  similar  kind 
of  glass. 

And  now  for  the  solution  of  the  problem  of  lighting  dark 
rooms  in  flats  and  apartments — lighting  them  by  day  as  well  as 
by  night — even  when  the  only  window  is  on  an  air-shaft.  The 
air-shaft  should  be  light  in  color,  white  or  ivory;  then  its  walls 
will  pick  up  and  transmit  the  maximum  of  light  from  above  and 
outside,  provided  there  is  a  lateral  opening.  The  only  thing  to 
do  in  lighting  a  room  thus  opening  on  an  air-shaft  is  to  make 
its  walls  and  ceiling  and  furniture  light  in  tone.  Then  in  the 
daytime  they  will  pick  up  from  the  air-shaft  all  the  light  there 
is  and  by  reflection  and  re-reflection  will  give  an  appearance  of 
cheerfulness  and  comfort  to  the  interior. 

One  is  often  amazed  to  learn  that  while  some  owners  of  apart- 
ments have  been  fully  instructed  on  this  point  they  fail  to  act. 
The  reason  is  as  follows,  as  I  learned  by  questioning  several 
janitors.  To  finish  the  wood-work  and  doors  of  a  small  room 
in  dark  tones  with  cheap  varnish  costs  about  75  cents  and  the 
varnish  lasts  on  an  average  about  three  years ;  so  that  the  real 
expense  per  room  is  only  25  cents.  To  finish  the  same  room  in 
white  or  ivory  enamel  costs  $2.00  and  the  ivory  enamel  often 
needs  renewing  every  year.  Consequently  the  apartment  owner 
balances  $2.00  in  one  hand  against  25  cents  in  the  other,  and  says 
he  will  stick  to  the  cheap  varnish. 

In  decorative  and  economical  lighting  it  is  best  to  favor  gold 
and  yellows  and  oranges  and  greens  of  light  tone,  at  the  expense 
of  reds  and  blues,  particularly  these  in  dark  tones.  Dark  red 
and  dark  green  or  blue  shades  on  lamps  or  fixtures  simply  kill 
the  light.  Especially  effective  in  toning  light  for  residence  use 
are  ground-glass  shades  of  various  shapes  and  sizes.  These  not 
only  eliminate  the  burning  effect  of  the  light  source,  but  they 
also  cream  the  white  slightly  by  eliminating  some  of  the  blue. 

I  dare  say  that  by  this  time  I  have  made  it  clear  why  I  regard 
the  problem  of  illuminating  residences  and  apartments  as  a  prob- 


home:  lighting  153 

lem  for  the  decorator  and  architect  rather  than  for  the  illuminating 
engineer.  But  I  want  to  say  right  here  that  unless  the  decorator 
and  the  architect  take  lessons  from  the  illuminating  engineer 
they  will  make  terrible  mistakes. 

DISCUSSION. 

Mr.  George  S.  Barrows:  Mr.  Hunter's  paper  presents  a  part 
of  the  subject  which  many,  who  are  qualified  to  solve  the  prob- 
lems ordinarily  presented,  may  find  unfamiliar;  and,  therefore, 
for  the  best  interest  of  the  ultimate  consumer,  they  should  prob- 
ably consult  with  some  one  familiar  with  that  side  of  the  question, 
which  they  have  not  studied  in  detail — the  architect  or  decorator. 

It  is  quite  likely  that  neither  the  architect,  the  decorator  nor 
the  illuminating  engineer  would  be  quite  satisfied  with  the  solu- 
tion of  the  problem  solely  by  the  other  one,  and  therefore  a 
compromise  must  be  effected  so  that  the  result  will  be  neither 
decorative  alone  on  the  one  hand  nor  utilitarian  alone  on  the 
other  hand. 

In  one  town  where  I  am  somewhat  familiar  with  the  method 
of  operation  of  the  illumination  experts  of  a  large  gas  company, 
the  salesman  endeavors  to  meet  the  architect  and  have  him 
design  the  fixtures,  suggesting,  however,  the  number  and  the 
location  of  the  outlets.  The  architect  also  indicates  the  design  of 
the  glassware  and  the  general  color  of  the  light  which  he  thinks 
should  be  used  to  best  emphasize  the  details  of  the  decoration.  It 
is  then  up  to  the  illumination  expert  to  see  that  the  proper 
burners,  mantles  and  glassware  are  used,  and  if,  in  his  opinion, 
the  requirements  of  the  architect  will  not  give  satisfactory  results, 
a  conference  is  arranged  in  order  to  discuss  and  eliminate  the 
points  of  difference. 

Mr.  G.  B.  Nichols  :  This  subject  is  one  of  very  vital  interest 
to  the  Illuminating  Engineering  Society  at  the  present  time,  for 
to  obtain  the  highest  efficiency  in  any  system  of  illumination  as 
planned  for  a  new  undertaking,  there  must  be  this  co-operation 
of  all  concerned  in  designing  the  entire  structure. 

In  general  the  owner  holds  off  in  making  up  his  mind  to  build 
until  the  last  moment  or  until  the  money  needed  has  been  entirely 
raised,  and  then  he  wishes  building  construction  to  start  at  once. 
This  gives  the  architect  but  very  little  time  to  prepare  his  plans 


154  TRANSACTIONS  I.  E.  S. — PART  II 

so  as  to  meet  the  owner's  wishes  or  so  that  building  con- 
struction can  be  started  at  the  proper  season  of  the  year.  After 
the  architect  has  the  plans  well  under  way  (in  most  cases 
about  eighty  to  ninety  per  cent,  completed)  he,  or  the  owner 
himself,  will  then  engage  specialists  to  co-operate  in  planning 
the  different  branches  of  engineering  work  required,  such 
as  illuminating,  electrical,  heating,  ventilating,  sanitary,  etc. 
The  architect  is  so  rushed  at  this  last  stage  of  the  work,  that 
minor  details  such  as  color  and  character  of  finish  of  walls  and 
illuminating  fixtures  are  wholly  deferred  until  after  the  building 
is  well  under  way  and  more  time  can  be  given  to  their  considera- 
tion. The  illuminating  engineer  therefore  must  in  most  cases 
plan  the  outlets  without  knowing  very  much  about  the  details 
of  the  rooms  aside  from  general  shape  and  use.  Furthermore, 
the  character  of  the  illuminating  fixtures  is  seldom  considered  at 
this  time,  the  illuminating  engineer  being  informed  to  provide 
enough  outlets  and  sufficient  wiring  to  take  care  of  any  system 
afterwards  required. 

A  great  deal  of  the  trouble  now  causing  poor  illumination 
can  be  traced  directly  to  the  fixture  houses  who  have  not  in  the 
least  tried  to  adapt  their  material  and  machinery  to  the  present 
standards  now  being  adopted,  but  rather  force  a  great  many  of 
their  antiquated  and  inefficient  fixture  designs  to  the  front  without 
any  regard  for  illumination,  or  to  the  expense  of  upkeep  of  them. 

In  summing  up  it  can  truly  be  said  that  the  architect  needs 
the  illuminating  engineer  to  guide  him  in  the  selection  of  the 
appropriate  illuminant;  the  illuminating  engineer  needs  the  archi- 
tect's advice  to  design  the  proper  standard  or  hanging  for  the^ 
illuminant  and  to  approve  the  location  and  type  of  same  so  as  to 
harmonize  with  the  architectural  design  of  the  building,  and  at 
the  same  time  the  architect  and  the  illuminating  engineer  need 
the  skill  of  the  decorator  to  blend  the  colors  and  wall  decora- 
tions so  as  to  add  to  the  dignity  and  efficiency  of  the  architect's 
design  and  the  illuminating  engineer's  conception  of  light  and 
shadows.  The  fixture  manufacturer  should  also  lend  his  aid  in 
giving  ideas  in  adapting  the  fixture  design  selected  to  the  most 
economical  method  of  manufacture  so  as  to  get  the  results  both 
in  respect  to  design  and  illumination  at  the  lowest  cost. 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 


Published  monthly,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIII 


APRIL,  1913 


No.  4 


Erratum. 

The  illustration  at  the  bottom  of 
page  opposite  page  106  of  March,  1913 
(Vol.  VIII,  No.  3)  Transactions  has 
the  words  "light"  and  "dense"  trans- 
posed. The  globe  on  the  left  should 
have  been  designated  as  dense,  and  the 
one  on  the  right  light. 


Council  Notes. 

The  April  meeting  of  the  council  was 
held  on  the  nth  instant,  in  the  general 
offices  of  the  society,  29  West  39th 
Street,  New  York.  Those  in  attendance 
were:  Preston  S.  Millar,  president; 
L.  B.  Marks,  treasurer;  C.  J.  Russell, 
George  S.  Barrows,  W.  J.  Serrill,  J.  D. 
Israel,  general  secretary;  V.  R.  Lau- 
singh,  Norman  Macbeth,  Alan  Bright 
(representative  of  Mr.  Howard  S. 
F,vans,  vice-president  of  the  Pittsburgh 
section)  and  James  T.  Maxwell  and 
H.  E.  Ives  by  invitation. 

A  monthly  report  on  the  finances  and 
membership  of  the  society  was  received 
from  the  assistant  secretary.  Accord- 
ing to  the  report  of  the  current  assets 
of  the  society,  as  of  April  1,  was 
$6,818;  $4,313  of  that  amount  repre- 
sented cash  in  bank;  the  remaining 
$2,505  represented  accounts  owing  to 
the  society.  The  unpaid  bills,  April  1, 
amounted  to  $775.  The  net  current 
assets     as     of     that     date,     therefore, 


amounted  to  $6,043.  The  expenses  for 
the  first  three  months  of  1913  was  said 
to  have  aggregated  $1,821.  The  mem- 
bership, including  the  additions  and  the 
defections  presented  at  the  meeting, 
totaled  1,338  members.  The  membership 
at  the  beginning  of  the  year  was  1,335 
members. 

Vouchers  Nos.  1241  to  1272,  inclusive, 
aggregating  $1,033.70,  which  had  been 
approved  and  submitted  by  the  finance 
committee,    were   authorized   paid. 

Mr.  L.  B.  Marks,  chairman  of  the 
committee  on  illumination  primer,  re- 
ported that  the  primer  was  to  be  trans- 
lated into  French  by  Professor  A. 
Blondel  of  Paris  and  published  within 
the  near  future.  It  was  understood  that 
the  translation  by  Professor  Blondel 
would  not  be  published  by  any  com- 
mercial organization.  It  was  resolved 
that  a  set  of  the  electros  of  the  illus- 
trations in  the  primer  be  sent  gratis  to 
Professor  Blondel  with  the  suggestion, 
that,  if  at  any  time  they  should  be 
utilized  for  publication  of  the  primer 
by  any  commercial  organization,  the 
society  would  desire  to  be  reimbursed 
for  their,  cost. 

The  executive  committee  was  em- 
powered to  undertake  preliminary  nego- 
tiations looking  toward  a  course  of 
lectures  on  architecture  and  decoration 
with  special  reference  to  illumination. 

A  monthly  report  on  activities  of  the 


TRANSACTIONS   I.    E.    S. — PART   I 


Philadelphia  section  was  received  from 
Mr.  W.  J.  Serrill,  vice-president. 

A  brief  report  on  the  work  of  the 
Chicago  section  was  received  from 
Vice-president  Cravath. 

President  Millar  announced  a  number 
of  additional  appointments  to  several 
committees.     These  were  approved. 

The  officers  of  the  society  were  em- 
powered to  appoint  from  time  to  time 
representatives  in  cities  not  having  sec- 
tions of  the  society. 

The  executive  committee  was  em- 
powered to  act  and  fix  a  time  and  place 
for  the  1913  convention  of  the  society 
upon  recommendation  of  the  committee 
on  time  and  place  for  the  annual  con- 
vention. 

A  list  of  members  who  were  delin- 
quent in  the  payment  of  their  1913  dues 
was  read. 


Section  Notes. 

CHICAGO  SECTION 

The  April  meeting  of  the  Chicago 
section  was  postponed.  Notices  of  the 
May  meeting  will  be  published  shortly. 

NEW    ENGLAND    SECTION 

The  New  England  section  held  a 
meeting  in  the  auditorium  of  the  Edison 
Building,  Boston,  April  21.  Dr.  Louis 
Bell  presented  a  paper  entitled  "Notes 
on  the  Effect  of  Radiation  on  the  Eye." 

NEW   YORK    SECTION 

At  a  meeting  of  the  New  York  sec- 
tion in  the  United  Engineering  Socie- 
ties' Building,  April  9,  Mr.  M.  Luckiesh 
of  the  National  Electric  Lamp  Associa- 
tion, Cleveland,  presented  a  paper  en- 
titled "Light  and  Art."  The  paper  was 
supplemented  by  a  series  of  demonstra- 
tions showing  the  effect  of  the  direction, 
the  quality  and  the  distribution  of  light 
on  various  art  objects.  Mr.  J.  B.  Taylor 
of  the  General  Electric  Company,  Sche- 


nectady, N.  Y.,  presented  a  paper  on 
"Color  Photography"  which  was  also 
supplemented  by  a  series  of  lantern 
slides  and  demonstrations.  About  135 
members  and  guests  were  present. 

PHILADELPHIA  SECTION 

The  Philadelphia  section  held  a  joint 
meeting  with  the  Philadelphia  Gas 
Works  section  of  the  National  Com- 
mercial Gas  Association  and  the  Phila- 
delphia Electric  Company  section  of  the 
National  Electric  Light  Association, 
April  23,  in  the  Parkway  Building, 
Broad  and  Cherry  Streets,  April  23. 
Mr.  T.  W.  Rolph  of  the  Holophane 
Works  of  the  General  Electric  Com- 
pany read  a  paper  on  "Metal  Reflectors 
for  Industrial  Lighting." 

PITTSBURGH   SECTION 

At  a  meeting  of  the  Pittsburgh  sec- 
tion, April  18,  in  the  Oliver  Building, 
Mr.  J.  Frank  Martin  of  the  Duquesne 
Lighting  Company  presented  a  paper 
entitled  "The  Illumination  of  Motion 
Picture  Projectors."  The  paper  appears 
in  this  issue  of  the  Transactions. 
About  30  members  were  present. 

A  paper  on  "Train  Lighting"  by  Mr. 
J.  L.  Minick  of  the  engineering  depart- 
ment of  the  Pennsylvania  Railroad  is 
scheduled  for  a  meeting  on  May  9. 


New  Members. 

The  following  twenty-four  applicants 
were  elected  members  of  the  society  at 
a  meeting  of  the  council,  April  11,  1913: 
Ashley,  Edward  E-,  Jr. 

Consulting     Engineer,     Starrett     & 
Van    Vleck,    45    East    17th    Street, 
New  York,  N.  Y. 
Austrom,  Charles  A. 

Assistant  Chief  Engineer,  The 
Travelers  Insurance  Co.r  Hartford, 
Conn. 


TRANSACTIONS    I.    E.    S. — PART    I 


Barrows,  Robert  Y. 

Chief  Designer,  The  Mitchell  Vance 
Company,  507  West  24th  Street, 
New  York,  N.  Y. 

Burrows,  W.  R. 

General  Electric  Co.,  5th  &  Sussex 
Streets,  Harrison,  N.  J. 

Coles,  J.  M. 

Mgr.  City  Dept.  and  Engineer,  Gen- 
eral Gas  Light  Co.,  80  Murray 
Street,   New  York,   N.  Y. 

Cox,  Edward  L. 

Secretary,  The  Enos  &  Watkins  Co., 
36  West  37th  Street,  New  York, 
N.  Y. 

Dibelius,  Ernest  F. 

The  New  York  Edison  Company, 
124  West  42nd  Street,  New  York, 
N.  Y. 

Deshon,  F.  B. 

Deshon-Davidson  Co.,  z22>  Main 
Street,   Tulsa,   Okla. 

Dawsen,  H.  E. 

Genera!  Electric  Company,  Harri- 
son, N.  J. 

Harrington,  R.  E. 

General  Electric  Company,  Harri- 
son, N.  J. 

Hewlett,  Arthur  T. 

Hewlett-Basing  Studio,  298  Fulton 
Street.   Brooklyn,   N.  Y. 

HlPPLE,   W.   C. 

Superintendent,  Westinghouse  Lamp 
Co.,  514  West  23rd  Street,  New 
York,  N.  Y. 

Ingraham,  Edgar  B. 

General    Electric    Co.,    30    Church 

Street,  New  York,  N.  Y. 
Kerr,  Thomas  T. 

The    New   York   Edison    Company, 

124  West  42nd   Street,   New  York, 

N.  Y. 

LePage,  Clifford  B. 

Stevens  Institute  of  Technology, 
Hoboken,  N.  J. 


Mullen,  Homer. 

General  Electric  Company,  Harri- 
son, N.  J. 

Oday,  A.  B. 

General  Electric  Company,  Harri- 
son, N.  J. 

Peck,  Robert  C. 

Electrical  Testing  Laboratories,  80th 
Street  &  East  End  Avenue,  New 
York,  N.  Y. 

Perry,  J.  W. 

General  Manager,  H.  W.  Johns- 
Manville  Company,  Madison  Avenue 
&  41st  Street,  New  York,  N.  Y. 

Shakin,   Victor. 

Electrical  Testing  Laboratories,  80th 
Street  &  East  End  Avenue,  New 
York,  N.  Y. 

Summers,  John  A. 

General  Electric  Company,  Harri- 
son, N.  J. 

Thistlewhite,  R. 

New  York  Electrical  School,  39 
West  17th  Street,  New  York,  N.  Y. 

Warner,  J.  Paul. 

Iron  City  Engineering  Co.,  1172 
Frick  Annex  Bldg.,  Pittsburgh,  Pa. 

Wyatt,  Chas.  K. 

Salesman,  H.  W.  Johns-Manville 
Company,  Madison  Avenue  &  41st 
Street,  New  York,  N.  Y. 


Sustaining  Members. 
At   a   meeting   of    the   council,    April 
11,   1913,  the  following  companies  were 
elected     sustaining     members     of     the 
society : 

Consolidated  Gas  Company  of  Boston. 
The   Benjamin    Electric   Company. 
The  Commonwealth  Edison  Company 

of  Chicago. 
The  Edison  Electric  Illuminating  Co. 

of  Brooklyn. 
National  Electric  Lamp  Association. 
The   Macbeth-Evans   Glass   Company. 
Westinghouse  Lamp   Company. 


TRANSACTIONS    I.    E.    S. — PART    I 


German  I.  E.  S. 

The  first  general  meeting  of  the  Ger- 
man Illuminating  Engineering  Society 
was  held  on  the  25th  of  February  in  the 
large  auditorium  of  the  Physical  Insti- 
tute of  the  University  of  Berlin. 

The  meeting  was  called  to  order  by 
the  chairman  of  the  temporary  council, 
Dr.  E.  Warburg,  president  of  the 
Physikalisch-technische  Reichsanstalt.  In 
his  address  President  Warburg  stated 
that  the  chief  aim  of  the  society  was 
to  attain  agreement  regarding  the  light 
unit,  regarding  nomenclature  and  re- 
garding methods  of  measurement,  and 
to  establish  standards.  He  emphasized 
very  strongly  the  necessity  for  inter- 
national agreement  in  these  matters  and 
pointed  out  the  great  significance  of 
such  agreements  for  the  development  of 
science  and  industry.  He  pointed  out 
the  function  of  the  society  in  uniting 
and  bringing  into  harmony,  theory  and 
practice  as  regards  illumination,  and 
stated  that  on  account  of  these  impor- 
tant factors,  the  Physikalisch-technische 
Reichsanstalt  had  done  its  utmost  to 
further  the  formation  of  the  society. 
Especially  interesting  was  a  statement 
that  at  the  Reichsanstalt.  experiments 
are  in  progress  from  which  they  have 
great  expectations  of  realizing  a  primary 
standard  of  light  which  will  receive 
international  sanction.  This  is  founded 
on  the  use  of  the  black  body  as  a  radia- 
tor, held  at  a  definite  temperature. 

The  secretary,  Dr.  E.  Liebenthal,  re- 
ported that  51  had  taken  part  in  the 
meeting  for  organization  in  November 
and  that  now  the  number  of  members 
had  grown  to  211.  The  committee  on 
permanent  officers  reported,  and  the 
following  were  elected :  president, 
Prof.  Dr.  E.  Warburg;  vice-president, 
Geheimrat  hummer ;  vice-president.  Ge- 


heimrat  Harber;  secretary.  Prof.  Lieben- 
thal ;  secretary,  Dr.  Kriiss ;  treasurer, 
Direktor  Schaller;  president  of  the 
council,  Geh.  Ober-Postrat  Dr.  Strecker; 
chairman  of  committee  on  light  unit, 
Geh.  Rat  Hagen;  chairman  of  com- 
mittee on  nomenclature,  Dr.  Strecker; 
chairman  of  committee  on  methods 
of  measurement,  Geh.  Rat  Brodhun. 

—  Tra  nslated from  Zeiisch  rift  fu  r 

Releuchtungswesen  by  C.  H.  Sharp. 


Glare  from  Reflecting  Surfaces. 

The  committee  on  glare  from  reflect- 
ing surfaces  is  preparing  an  eight-page 
pamphlet  which  will  bring  to  the  atten- 
tion of  the  reader  the  necessity  of 
eliminating  glare  from  glazed  paper, 
polished  or  glass  desk  tops,  glazed 
blackboards  and  waUs.  Two  leaves  of 
the  pamphlet  will  be  of  highly  calen- 
dered stock,  while  the  two  remaining 
leaves  will  be  of  unglazed  book  paper. 
It  is  hoped  that  this  pamphlet  will  im- 
press upon  each  reader  the  importance 
of  eliminating  from  general  use  all 
polished  surfaces  and  especially  glazed 
paper.  As  paper  is  the  most  common 
source  of  glare  with  which  the  com- 
mittee is  concerned,  data  are  constantly 
being  collected  on  printing  processes 
and  available  matt-surface  papers.  This 
enables  the  committee  to  co-operate  with 
publishers  who  desire  to  enlist  in  the 
army  of  vision  conservationists. 


Factory  Lighting  Legislation. 

Quoted  below  is  a  section  of  Bill 
No.  26  of  the  laws  of  the  State  of 
New  York  entitled  "An  Act  to  amend 
the  labor  law,  in  relation  to  the  pro- 
tection of  employees  operating  machin- 
ery, dust  creating  machinery,  and  the 
lighting   of    factories   and   workrooms.'' 


TRANSACTIONS    I.    E.    S.— PART    I 


This  section  of  the  bill,  which  relates 
particularly  to  lighting  of  factories, 
passageways  and  workrooms,  was 
drafted  in  accordance  with  recommen- 
dations made  by  the  committee  on  fac- 
tory lighting  legislation  of  the  Illumi- 
nating Engineering  Society.  The  bill 
was   signed   by   Governor    Sulzer   April 

i/.   IQI3- 

All  passageways  and  other  portions  of  a  factory, 
and  all  moving  parts  of  machinery  which  are  not 
so  guarded  as  to  prevent  accidents,  where,  on  or 
about  which  persons  work  or  pass  or  may  have  to 
work  or  pass  in  emergencies,  shall  be  kept  prop- 
erly and  sufficiently  lighted  during  working 
hours.  The  halls  and  stairs  leading  to  the  work- 
rooms shall  be  properly  and  adequately  lighted, 
and  a  proper  and  adequate  light  shall  be  kept 
burning  by  the  owner  or  lessee  in  the  public 
hallways  near  the  stairs,  upon  the  entrance  door 


and  upon  the  other  floors  on  every  work  day  in 
the  year,  from  the  time  when  the  building  is 
open  for  use  in  the  morning  until  the  time  it  is 
closed  in  the  evening,  except  at  times  when  the 
influx  of  natural  light  shall  make  artificial  light 
unnecessary.  Such  lights  shall  be  so  arranged  as 
to  insure  their  reliable  operation  when  through 
accident  or  other  cause  the  regular  factory  light- 
ing is  extinguished. 

All  workrooms  shall  be  properly  and  adequate- 
ly lighted  during  working  hours.  Artificial  il- 
luminants  in  every  workroom  shall  be  installed, 
arranged  and  used  so  that  the  light  furnished 
will  at  all  times  be  sufficient  and  adequate  for 
work  carried  on  therein,  and  so  as  to  prevent 
unnecessary  strain  on  the  vision  or  glare  in  the 
eyes  of  the  workers.  The  industrial  board  may 
make  rules  and  regulations  to  provide  for  ade- 
quate and  sufficient  natural  and  artificial  light- 
ing facilities  in  all  factories. 

This  act  shall  take  effect  October  first,  nine- 
teen hundred  and  thirteen. 


0- 


TRANSACTIONS 

OF  THE 

Illuminating 
Engineering  Society 

APRIL,  1913 
PART  II 

Papers,  Discussions  and  Reports 


[  APRIL,  1913  ] 

CONTENTS  --  PART  II 


A  Photometer  Screen  for  Use  in  Tests  of  Street  Illumina- 
tion.    By  Arthur  H.  Ford 155 

The  Flame  Carbon  Arc  Lamp.     By  W.  A.  Darrah 162 

The   Illumination   of   Motion   Picture   Projectors.      By  J. 
Frank  Martin 180 


/&V 


A  PHOTOMETER  SCREEN  FOR  USE  IN  TESTS  OF 
STREET  ILLUMINATION. 


BY  ARTHUR   H.   FORD. 


Synopsis:  This  paper  puts  forth  the  proposal  that  in  illumination 
measurements  a  test-plate  with  rounded  surfaces  or  one  with  several  plane 
surfaces,  the  mean  illumination  of  which  has  been  determined,  be  used 
in  place  of  the  usual  flat  translucent  photometer  test-plate.  It  gives 
comparative  illumination  readings  obtained  with  the  use  of  five  different 
test-plates. 

The  purpose  of  street  illumination  being  the  making  visible 
of  obstructions  to  traffic,  which  obstructions  seldom  have  plane 
surfaces  making  definite  angles  with  the  street  surface,  the 
ordinary  method  of  measuring  street  illumination  on  a  horizontal 
plane  or  one  normal  to  the  ray  of  light  does  not  give  the  infor- 
mation desired.  The  writer  proposes  to  overcome  this  defect 
by  using,  as  a  screen,  a  body  with  rounded  surfaces  or  several 
plane  surfaces;  the  mean  illumination  of  which  is  determined. 
Such  a  screen  would  correspond  to  a  brick  if  plane  surfaces  were 
used  or  a  stone  if  rounded  surfaces  were  used ;  and  the  illumina- 
tion would  be,  to  a  considerable  extent,  independent  of  the 
direction  from  which  the  light  comes. 

This  paper  is  the  record  of  some  tests  with  screens  having 
various  configurations. 

Since  the  surroundings  of  the  photometer  cannot  be  controlled 
in  street  photometry  as  in  laboratory  work,  it  is  obvious  that  the 
surfaces  used  must  be  those  of  a  translucent  body  attached  to 
the  viewing  part  of  the  photometer.  The  screens  tested  were 
mounted,  in  turn,  on  the  elbow  tube  of  a  Sharp-Millar  pho- 
tometer and  readings  of  the  instrument 'made  as  the  angular 
position  of  the  test  lamp  was  changed  through  90  deg.  Polar 
curves  were  then  plotted  between  the  angular  position  of  the 
test  lamp  and  readings  of  the  photometer,  in  terms  of  the 
maximum  reading  obtained  with  the  particular  screen  being 
tested. 


i56 


TRANSACTIONS  I.  E.  S. — PART  II 


The  following  screens  were  used: 

No.  i. — A  plate  of  translucent,  milk  white  glass  with  a  ground 
surface  (Fig.  i),  the  glass  being  the  regular  diffusing  screen  fur- 
nished with  the  photometer. 

No.  2. — A  cube  of  paraffin  i^  inch  (4.1  cm.)  on  a  side;  one 
side  covered  with  an  opaque  screen  (Fig.  2).  While  paraffin 
is  not  a  suitable  substance  for  making  permanent  screens,  the 
ease  of  moulding  it  into  irregular  shapes  adapts  it  admirably  for 
making  screens  for  temporary  use.  Its  optical  properties  are 
satisfactory  if  the  screens  are  carefully  selected  for  uniformity 
of  texture  and  optical  symmetry. 

No.  3. — One  quarter  of  a  paraffin  sphere  1^  inch  (4.1  cm.) 
in  radius  having  one  flat  side  covered  with  an  opaque  screen 
(Fig.  3)- 

No.  4. — An  "Alba"  glass  hemisphere  3  inches  (7.6  cm.)  in 
diameter  having  one  half  blackened  Fig.  4).  This  globe  is  about 
yi  inch  (0.3  cm.)  thick;  the  translucence  being  due- to  air  bubbles 
in  the  glass. 

No.  4a. — The  same  screen  as  No.  4,  but  with  the  addition  of 
a  piece  of  paper  on  the  flat  side  of  the  hemisphere. 

The  following  data  were  obtained: 
Screen  No.  i. 


Angle  of 
lamp. 

Photometer 
reading. 

90 
80 

IOO 

97 

70 
60 

90 
81 

50 
40 
30 

70 
60 
46 

20 

34      . 

IO 

— 

O 

— 

Screen  N°  I. 


ford:     a  photometer  SCREEN 


157 


Screen  No.  2. 


Angle  of 
lamp. 

90 

80 

70 

60 

50 

40 

30 

20 

IO 

O 


Angle  of 
lamp. 

90 

80 

70 

60 

50 

40 

30 

20 

IO 

o 


Angle  of 
lamp. 

90 


70 
60 

50 
40 
30 
20 
10 
o 


Photometer 
reading. 

57 

65 

81 

85 
100 
100 

95 
96 

99 

75 


®IOO 


Photometer 
reading. 

75 

87 

94 

97 
100 
100 

93 

9i 

84 

66 


Screen  No.  3 
°\0° 


Screen  No.  4. 


Screen  Ns  3. 

Fig-  3- 


Photometer 
reading. 

IOO 
70 
52 
46 
42 
37 
35 
33 
28 
22 


158 


TRANSACTIONS  I.  E.  S. — PART  II 


Screen  No.  4a. 


Angle  of 
lamp. 

90 

80 

70 

60 

50 

40 

SO 

20 

IO 

o 


°»0°      AftO 


Photometer 
reading. 

70 

82 

92 
IOO 
IOO 

97 

9i 

87 

72 

61 


Screen  N°4A. 


Fig.  4a. 

CONCLUSION. 

Screens  Nos.  3  and  4a  are  satisfactory  for  use  in  street  pho- 
tometry, with  the  advantage  somewhat  in  favor  of  No.  3  because 
of  the  small  dependence  of  illumination  determinations  on  the 
direction  of  the  light  source.  The  writer  would  suggest  the  use 
of  a  screen  of  this  form  made  of  "Alba"  or  some  similar  glass. 


DISCUSSION. 

Dr.  C.  H.  Sharp  (communicated)  :  Prof.  Ford  makes  the 
statement  that  the  ordinary  method  of  measuring  street  illumina- 
tion on  a  horizontal  plane  or  on  one  normal  to  the  ray  of  light, 
does  not  give  all  the  information  requisite  to  determine  the 
visibility  of  objects  on  the  surface  of  the  street.  This  is  quite 
true.  He  proposes  to  overcome  this  difficulty  by  using  a  test  screen 
which  gives  a  value  which  represents  an  indeterminate  mixture  of 
illumination  values  in  various  planes.  It  is  not  at  all  clear  that 
this  represents  a  solution  of  the  difficulty. 

In  the  first  place,  it  must  be  noted,  as  Millar  has  pointed  out, 
that  objects  on  the  street  are  seen  because  of  a  difference  of  the 
brightness  of  the  object  and  its  background.  A  brick  lying  on  the 
street  may  be  seen  because  it  is  whiter  than  the  street  surface ;  or 
because  to  the  observer  a  vertical  surface  of  the  brick  is  brighter 
than  the  horizontal  surface  of  the  pavement,  a  condition  which 
may  be  due  either  to  the  brick  being  actually  whiter  or  to  the 


A   PHOTOMETER  SCREEN  159 

illumination  on  the  vertical  plane  being  greater  than  that  on  the 
horizontal  plane ;  or  it  may  be  seen  because  a  vertical  surface  of 
the  brick  is  darker  than  the  street  surface,  which  may  be  because 
it  lies  in  its  own  shadow.  There  are  then  various  conditions 
which  determine  visibility  and  which  arise  from  the  relations  be- 
tween the  object  viewed,  the  lamps  illuminating  it,  and  the  rela- 
tive reflecting  power  of  the  object  and  of  the  street  surface,  as 
well  as  upon  the  illumination  in  various  planes.  Therefore,  even 
if  we  know  all  there  is  to  be  known  about  the  illumination  in  all 
the  various  planes  conceivable,  we  could  not  say  that  any  brick 
would  be  visible  from  any  point  unless  we  had  some  other  items 
of  information  besides.  Illumination  values  alone  would  not  tell 
us.     Therefore  at  best  Prof.  Ford's  solution  is  an  imperfect  one. 

From  a  scientific  point  of  view  also  his  proposition  seems  to 
be  untenable.  When  we  consider  the  illumination  at  a  certain 
point  on  a  certain  plane,  we  are  dealing  with  a  perfectly  definite 
physical  quantity.  The  only  arbitrary  question  is  how  perfectly 
the  test  plate  used  in  making  the  measurement  of  that  illumination 
conforms  to  the  ideal  law  of  diffusion.  When,  however,  we 
measure  the  brightness  of  a  certain  piece  of  parafhne  or  a  certain 
hemisphere  of  a  certain  kind  of  glass,  we  have  a  result  which  is 
purely  arbitrary  and  incapable  of  interpretation  in  simple  terms. 
It  is  also  very  questionable  whether  hemispheres  of  diffusing  glass 
could  under  the  best  conditions  be  made  so  like  each  other  that 
they  would  always  show  the  same  diffusion  characteristics. 
Unless  it  were  possible  to  reproduce  these  hemispheres,  it  would 
be  impossible  to  make  all  illumination  photometers  give  the  same 
results  under  the  same  conditions.  And  there  is  no  reason  to 
suppose  that  in  the  present  state  of  the  art  of  glass  making,  this 
can  be  done. 

It  would  be  interesting  if  Prof.  Ford  were  to  study  the  illumina- 
tion on  a  street  using  his  proposed  test  plate,  and  were  then  to 
present  a  paper  to  the  society  telling  just  what  that  illumination 
was  in  actual  physical  units. 

Mr.  M.  Luckiesh  (communicated)  :  Prof.  Ford  presents 
some  interesting  data  regarding  the  problem  of  measuring 
"illuminating  efficiency."  It  is  true  that  street  lighting  presents 
some    distinct    problems    in    the    matter    of    rating    illumination 

2 


160  transactions  i.  e.  s. — part  ii 

according  to  its  ability  to  make  objects  visible  but  interior  lighting 
is  not  free  from  the  same  problems.  Various  methods  for 
measuring  "illuminating  efficiency"  have  been  suggested  but  the 
one  used  at  present,  although  deficient,  will  no  doubt  be  adhered 
to  until  illuminating  engineers  agree  on  the  answer  to  the  ques- 
tion. How  does  the  direction  of  the  incident  light  affect  the 
ability  of  that  light  to  make  objects  visible? 

This  question  will  not  be  answered  without  fully  analyzing 
illumination.  With  this  in  mind  the  writer  has  studied  in  detail 
the  distribution  of  natural  and  artificial  light  in  interiors.  This 
work  which  appeared  in  the  Transactions  for  October,  1912,  is 
only  a  beginning  of  the  analysis  which  must  be  made  in  order  to 
conclude  just  how  lumens  incident  from  various  directions  should 
be  weighted. 

Prof.  Ford's  screens,  Nos.  3  and  4a,  which  he  considers  satis- 
factory for  use  in  street  photometry  take  into  account  light 
incident  within  an  angular  range  of  90  deg.  (considering  only  the 
vertical  plane  parallel  to  the  direction  of  the  street).  Obviously 
the  screen  or  the  whole  photometer  must  be  rotated  180  deg.  in 
order  to  account  for  the  light  from  the  other  light  source  because 
there  would  be  positions  where  sources  on  opposite  sides  of  the 
photometer  would  contribute  light.  Just  how  he  would  weight 
these  two  measurements  is  not  stated. 

To  fully  emphasize  the  complexity  of  the  problem,  consider  a 
case  where  one  sees  a  vehicle  in  silhouette  against  a  bright  back- 
ground such  as  a  highly  illuminated  pavement.  In  such  cases 
visibility  depends  upon  the  contrast  of  the  dark  background; 
therefore  the  less  light  that  is  incident  on  the  vehicle  from  the 
general  direction  of  the  observer  the  more  visible  will  the 
object  be. 

These  points  are  cited  merely  to  emphasize  the  fact  that  more 
work  must  be  done  on  the  analysis  of  illumination.  In  view  of 
the  lack  of  agreement  regarding  the  method  of  determining 
"illuminating  efficiency",  the  writer  cannot  wholly  agree  with 
Prof.  Ford's  conclusion  that  screens  Nos.  3  and  4a  are  satis- 
factory for  street  illumination.  It  would  be  interesting  to  learn 
why  he  reached  this  conclusion.  Prof.  Ford's  data  are  very  in- 
teresting and  certain  of  those  screens  may  perhaps  find  immediate 


A    PHOTOMETER    SCREEN  l6l 

application  in  special  cases.  However,  the  data  will  be  of  greater 
use  after  illuminating  engineers  reach  an  agreement  and  decide 
what  they  desire  to  measure. 

Mr.  G.  H.  Sticknev  (communicated)  :  The  failure  of  the 
normal  intensity  or  that  on  a  horizontal  surface  to  express  the 
true  values  of  street  illumination  has  long  been  evident  and  not 
infrequently  has  been  the  subject  of  discussion  among  illuminat- 
ing engineers.  While  I  believe  it  is  generally  conceded  that  the 
true  values  lie  somewhere  between  those  determined  by  the  above 
mentioned  methods,  there  seems  to  be  no  general  agreement  as  to 
just  where  these  values  fall,  and  it  is  quite  probable  that,  under 
different  conditions,  their  relative  position  between  these  two 
extremes  may  vary. 

About  1906  Mr.  \Y.  D'A.  Ryan  and  myself  experimented  a 
little  with  a  translucent  hemisphere  and  also  with  other  spherical 
sections  partially  blackened  or  coated  with  tin  foil,  but,  on 
account  of  various  difficulties,  did  not  arrive  at  a  form  of  screen 
which  we  were  willing  to  recommend. 

Beyond  the  difficulty  of  determining  the  intermediate  values 
which  would  be  acceptable  as  fair  to  both  large  and  small  units 
with  wide  and  narrow  spacing,  we  were  unable  to  secure  a  screen 
which  would  make  accurate  measurements  on  very  low  intensities. 
It  will  be  remembered,  in  this  connection,  that  the  class  of 
problems  where  this  type  of  measurement  is  important  are  those 
in  which  the  actual  normal  intensities  may  run  as  low  as  1/100 
foot-candle  or  less.  While  we  obtained  accurate  settings  with 
diffusing  screens  of  opal  glass  and  paraffine  with  values  above  one 
foot-candle,  we  did  not  have  much  success  with  low  intensities; 
that  is,  1/10  foot-candle  or  less. 

In  looking  over  the  curves  presented  by  the  author  I  would 
agree  that  screens  No.  3  and  4-a  show  a  better  characteristic 
curve  than  the  other  forms.  Still  I  believe  that  many  observers 
will  feel  that  both  of  these  types  give  too  much  weight  to  light  in 
angles  approximating  zero  degrees. 

\\  hile  at  present  the  difficulties  seem  almost  unsurmountable,  it 
must  be  recognized  that  there  is  an  insistent  demand  for  a  better 
method  of  determining  the  value  of  street  lighting  intensities  and 
any  work  tending  toward  this  end  will  be  of  great  value. 


l62  TRANSACTIONS  I.  E.  S. — PART  II 

THE  FLAME  CARBON  ARC  LAMP.* 


BY   W.   A.   DARRAH. 


Synopsis:  The  operating  characteristics  of  the  recently  developed 
flame  carbon  arc  lamp  upon  direct  and  upon  alternating  current  supply 
are  outlined  and  illustrated  in  the  following  paper.  Flame  carbons  and 
their  light  giving  properties  and  characteristics  are  also  discussed.  The 
author  states  that  this  lamp  is  peculiarly  well  adapted  to  street  and  large 
area  lighting  on  account  of  the  color  (which  approximates  a  daylight 
value  when  white  flame  carbons  are  used)  and  distribution  of  its  illumi- 
nation, and  on  account  of  its  comparatively  low  operating  and  main- 
tenance costs.  The  efficiency  of  this  type  of  lamp  (with  white  flame 
carbons)  quoted  in  mean  lower  hemispherical  candles  per  watt  is  said  to 
vary  from  2.5  with  light  opal  globes  to  5  with  clear  glass  globes ;  yellow 
carbons,  the  author  adds,  may  be  obtained  which  will  give  from  30  to 
50  per  cent,  more  light  for  the  same  energy  consumption. 

The  methods  of  lighting  to-day  are  being  revolutionized. 
Hundreds  of  thousands  of  dollars  worth  of  old  equipment,  is 
yearly  being  sent  to  the  scrap  pile,  to  be  replaced  by  more 
efficient,  more  effective,  more  economical  apparatus.  The  num- 
ber of  "great  white  ways"  is  being  multiplied  monthly.  Civic 
bodies  and  merchant's  associations  are  clamoring  for  more  light. 

One  of  the  fundamental  causes  of  the  transition  now  in  pro- 
gress, is  the  rapid,  and  very  complete  development  and  com- 
mercialization of  the  long-burning,  flame  carbon  arc  lamp.  The 
flame  carbon  arc  lamp  is  not  new.  It  does  not  even  in  its  present 
highly  developed  form,  depend  for  its  operation  upon  new 
theories,  or  recent  discoveries.  But  it  does  depend  for  its  present 
swift  and  complete  development  upon  the  awakening  of  civic 
pride,  the  discovery  of  the  enormous  commercial  value  of  light, 
and  the  accompanying  prosperity  which  has  swept  across  the 
country. 

The  subject  of  long  burning  flame  carbon  arc  lamps  is  too  large 
a  one  to  be  treated  in  a  limited  paper  of  this  kind.  It  is  proposed 
here  to  very  briefly  touch  -on  some  characteristics  of  long  burning 
flame  carbon  arc  lamps,  to  show  how  these  characteristics  particu- 
larly adapt  this  illuminant  for  some  fields  of  work,  and  to  then 

*A  paper  read  before  a  meeting  of  the  New  England  section  of  the  Illuminating 
Engineering  Society,  February  17,  1913. 


DARRAH  :      THE  FLAME  CARBON  ARC  LAMP  1 63 

consider  some  points  in  connection  with  the  design  of  these  lamps 
from  the  standpoint  of  the  operating  man. 

The  long  burning  flame  carbon  arc  lamp,  is  to-day,  without 
doubt,  the  most  efficient,  commercial  source  of  light  considering 
at  the  same  time  color  value;  and  what  is  unfortunately  not 
always  true  of  illuminants  of  this  class,  it  is  also  the  most 
economical  to  operate,  considering  the  average  conditions  of  cost 
of  energy,  labor,  and  materials  when  the  mean  lower  hemispheri- 
cal candle-power  required  exceeds  five  or  six  hundred. 

The  light  of  the  flame  carbon  lamp  is  steady  and  the  dis- 
tribution excellent  for  lighting  large  areas  with  considerable 
uniformity.  The  color  may  be  varied  within  a  rather  wide  range 
from  white  to  yellow  depending  upon  the  kind  of  carbons  em- 
ployed. Other  colors  such  as  red,  blue,  green,  etc.,  may  of  course 
be  obtained  if  sufficient  demand  for  them  should  arise,  as,  for 
instance,  for  advertising  purposes  or  where  special  colors  lend 
themselves  more  readily  for  special  work. 

A  further  highly  desirable  attribute  of  the  flame  arc,  and  one 
which  has  perhaps  not  yet  been  fully  appreciated  is  the  reduced 
intrinsic  brilliancy  over  the  carbon  arc,  due  to  the  relatively  large 
area  of  the  light  source.  While  the  intensity  of  the  flame  arc  is 
so  great  that  even  with  the  much  larger  area  of  the  light  source 
it  is  usually  desirable  to  use  diffusing  glassware,  yet  the  shadows 
are  far  softer  and  the  distribution  more  uniform  than  can  be 
obtained  with  the  older  forms  of  arcs. 

CHARACTERISTICS  OF  THE  FLAME  CARBON  ARC. 

A  consideration  of  the  flame  carbon  arc  is  only  of  value  in  con- 
nection with  a  lamp  and  a  specific  variety  of  carbons.  In  the 
statements  which  follow  an  enclosed  flame  carbon  arc  lamp 
similar  to  that  shown  in  Fig.  1  is  considered,  although  the  dis- 
cussion in  general  applies  to  other  lamps  of  this  type.  The  car- 
bons under  consideration  are  assumed  to  be  commercial  carbons, 
varying  in  diameter  from  V$  in.  to  %  in. 

In  appearance,  the  flame  carbon  arc  more  nearly  resembles  the 
metallic  flame  arc  than  any  other  type.  It  consists  of  a  long  ill- 
defined  flame  which  is  intensely  luminous  and  is  terminated  with 
a  bright  point  at  each  end.     Both  the  alternating  and  direct  cur- 


164  TRANSACTIONS  I.  E.  S. — PART  II 

rent  arcs  have  these  characteristics,  and  it  is  very  difficult  to  dis- 
tinguish the  direct  current  from  the  alternating  current  arc  by 
inspection.  No  crater  is  formed  on  the  positive  carbon  and  the 
distinct  cathode  spot  on  the  negative  carbon  is  less  conspicuous. 
It  is  probable  that  the  length  and  density  of  the  arc  material 
prevents  the  scouring  action  of  the  particles  emitted  by  the  arc, 
which  probably  cause  the  crater  in  the  case  of  the  open  direct  cur- 
rent arc.  Unlike  the  enclosed  carbon  arc,  the  flame  is  intensely 
luminous  and  is  the  source  of  a  considerable  amount  of  radiation 
in  the  red  and  green  portions  of  the  spectrum. 

The  majority  of  light  comes  from  the  luminous  flame,  although 
the  terminals  of  the  arc  have  a  higher  intrinsic  brilliancy  than  the 
other  portions. 

The  arc  may  have  any  color,  depending  upon  the  materials  used 
in  the  carbons.  Commercially,  white  and  yellow  light  carbons 
are  mainly  employed,  although  under  special  conditions,  for 
special  purposes,  other  colors  are  occasionally  used.  The  color 
of  the  arc  is  dependent  upon  the  current  density  to  some  extent. 
In  other  words,  an  arc  that  would  be  white  with  certain  values  of 
current  density,  may  become  distinctly  yellow  when  the  current 
density  is  materially  decreased.  It  appears  that  the  temperature 
of  the  arc  also  has  considerable  effect  upon  the  color,  since  an  arc 
which  will  burn  white  under  normal  conditions,  may  become  in- 
tensely yellow  when  the  temperature  is  lowered,  either  by  cooling 
the  terminals,  by  special  air  draft,  or  by  forming  the  arc  in  an 
atmosphere  which  will  readily  conduct  the  heat  away.  It  seems 
probable  that  the  change  in  color  may  be  produced  by  lowering 
the  temperature  of  the  minute  particles  which  are  luminous,  to 
a  yellow  heat  instead  of  allowing  them  to  remain  at  a  tempera- 
ture sufficiently  high  for  the  radiation  of  white  light. 

It  is  interesting  to  note  that  the  addition  of  water  vapor  to  the 
arc  chamber  will  cause  an  arc  normally  emitting  a  white  light  to 
burn  yellow.  A  further  suggestive  fact  is  that  certain  gases  may 
be  present  in  the  arc  chamber  which  will  cause  a  yellow  arc  to* 
become  intensely  white. 

Fig.  2  shows  an  alternating  current  flame  carbon  arc  between 
Y$  in.  carbons  and  an  atmosphere  of  carbon  monoxid  and  nitrogen. 
This  was  a  10-ampere  arc  with  a  potential  between  the  electrodes; 


<r 


ih 


Fig.  i.— Enclosed  flame  carbon  arc  lamp. 


Fig.  2. — A  io-ampere,  50-volt  alternating  current  flame  carbon  arc  between  %  in. 

carbon  and  an  atmosphere  of  carbon  monoxid  and  nitrogen.     The 

magnetic  blow  away  from  the  side  rod  is  clearly  shown. 


-A  6.5-ampere,  70-volt  direct  current  flame  arc  between  %  in. 
carbons  under  normal  conditions. 


DARRAH  :      THE  FLAME  CARBON  ARC  LAMP  165 

of  approximately  50  volts.  In  this  arc,  the  magnetic  blow  away 
from  the  side  rod  may  be  distinctly  noted. 

Figs.  3  and  4  show  direct  current,  6.5-ampere,  70-volt  arcs  be- 
tween %  in.  carbons.  It  will  be  noted  that  both  ends  of  the 
direct  current  arc  are  distinctly  luminous  and  very  similar  in 
appearance  to  the  alternating  arc.  Fig.  3  shows  the  arc  under 
normal  conditions,  while  Fig.  4  shows  the  so-called  non-magnetic 
arc  which  is  considerably  diffused  and  very  slightly  affected  by  a 
magnetic  field.  Figs.  5  and  6  illustrate  an  enclosed  carbon  arc 
between  unimpregnated  y2  in.  carbons  operating  under  similar 
conditions  to  the  flame  carbon  arc.  It  will  be  noted  here  that  the 
electrodes  of  the  enclosed  carbon  arc  are  intensely  bright.  Since 
the  arc  between  the  unimpregnated  carbons  emits  a  large  per- 
centage of  blue  and  ultra-violet  rays,  the  photograph  shows  this 
arc  considerably  more  luminous  than  it  appears  to  the  eye. 

The  diameter  of  the  flame  carbon  arc  under  normal  conditions 
is  approximately  Y%  in.  Curve  1,  of  Fig.  7,  shows  the  length  of 
the  flame  carbon  arc  with  various  voltages  applied  to  the 
terminals  and  with  constant  current.  It  will  be  noted  that  the 
length  of  the  arc  ranges  from  )A  in.  with  40  volts,  to  approxi- 
mately 5  in.  with  220  volts.  The  increase  is  substantially  a 
straight  line.  If  continued,  the  curve  would  pass  through  the  zero 
separation  point  with  about  30  volts  drop,  indicating  that  this  is 
the  loss  at  the  terminals  independent  of  the  length  of  the  arc. 
Calculated  on  this  basis,  an  arc  5  in.  in  length  and  having  a  total 
drop  of  220  volts  would  have  190  volts  loss  throughout  its  length, 
or  a  drop  of  38  volts  per  inch.  This  is  equivalent  to  about  3.8 
ohms  per  inch  of  length  for  a  10-ampere,  direct  current  arc,  which 
while  somewhat  higher,  is  not  very  materially  different  from  the 
conditions  often  obtained  in  the  enclosed  carbon  arc. 

Curve  2,  on  Fig.  7,  is  the  well-known  characteristic  curve  of 
an  arc.  It  will  be  noted  that  when  the  current  exceeds  approxi- 
mately 6.5  amperes,  the  flame  carbon  arc' becomes  exceedingly 
stable.  In  other  words,  the  drop  in  voltage  across  the  arc  in- 
creases in  approximately  the  same  ratio  that  the  current  is  in- 
creased. This  is  somewhat  different  from  the  titanium  and 
enclosed  arcs  which  are  correspondingly  very  much  more  un- 
stable.    The  causes  of  the  stability  of  the  flame  carbon  arc  on 


l66  TRANSACTIONS  I.  E.  S. — PART  II 

currents  above  6.5  amperes  are  mainly  two :  first,  the  relatively 
high  resistance  of  the  arc  and  the  increase  of  this  resistance  as 
the  length  increases ;  and,  second,  the  lower  evaporation  point  and 
greater  volume  of  the  materials  in  the  flame  carbon  arc.  In  this 
connection,  it  is  interesting  to  note  that  these  two  conditions,  to- 
gether with  the  higher  margin  between  the  temperature  at  which 
the  arc  is  a  conductor  and  the  tempearture  at  which  it  is  not,  allow 
of  the  satisfactory  operation  of  the  flame  carbon  arc  upon  cir- 
cuits, the  frequencies  of  which  is  as  low  as  25  cycles. 

Figs.  8,  9  and  10  respectively,  show  oscillograms  of  the 
titanium  arc,  the  enclosed  carbon  arc  and  the  flame  carbon  arc. 
It  will  be  noted  that  while  the  current  wave  has  substantially  the 
same  distortion  in  the  case  of  the  titanium  and  enclosed  carbon 
arcs,  it  is  very  much  more  nearly  a  sine  wave  in  the  case  of  the 
flame  arc.  An  inspection  of  the  oscillographs  shows  that  the  dis- 
tortion of  the  voltage  wave  is  very  considerably  greater  than  the 
distortion  of  the  current  wave.  The  titanium  arc  shows  a  dis- 
tinct peak  at  the  instant  at  which  the  current  is  interrupted  and  a 
very  steep  wave  front.  This  distortion  of  the  current  and  volt- 
age wave  in  the  titanium  arc  is  so  great  and  their  centers  are 
displaced  to  such  a  degree  that  the  resultant  power  factor  is 
approximately  50  per  cent. 

The  distortion  of  the  enclosed  carbon  arc  is  somewhat  less,  but 
the  displacement  of  the  current  and  voltage  waves  is  sufficient 
to  reduce  the  power  factor  to  approximately  80  per  cent.  It  is 
interesting  to  note  the  higher  harmonics  which  are  present  in 
the  voltage  wave  of  the  enclosed  carbon  arc  are  absent  in  both 
of  the  flame  arcs. 

The  voltage  wave  of  the  flame  carbon  arc  shows  less  distortion 
than  either  of  the  other  arcs,  and  the  displacement  is  so  slight 
that  the  power  factor  ranges  from  85  to  90  per  cent.  The  above 
discussion  should  be  understood  to  apply  entirely  to  the  power 
factor  of  the  arc  itself  and  to  be  independent  of  the  lamp 
mechanism  or  coils. 

The  advantages  of  the  higher  power  factor  are  well  known, 
while  the  advantages  of  the  smoother  wave  form  are :  lower 
voltage  strains,  more  stable  arcs,  and  less  induction  between  arc 
circuits  and  adjacent  telephone  or  telegraph  circuits. 


4- — A  6.5  ampere  70-volt  direct  current  flame  arc  between  %  in.  carbons.     The 
so-called  non-magnetic  arc  which  is  said  to  be  only  slightly  affected  by  a 
magnetic  field,  and  considerably  diffused,  is  plainly  illustrated. 


F»g-  5.— An   enclosed  carbon  arc  between 
' :  inch  uniinpregnated  carbons. 


Fig.  6.  — An  enclosed 
carbon  arc. 


Fig.  7. —Characteristics  of  flame  carbon  arc  lamps. 


Fig.  S. — Oscillograms  of  titanium  arc. 


Fig.  9. — Oscillograms  of  enclosed  carbon  arc. 


DARRAII  :      THE  FLAME  CARBON  ARC  LAMP 


l67 


As  previously  pointed  out,  the  smaller  distortion  and  higher 
power  factor  of  the  flame  carbon  arc  results  largely  from  the 
sloping  characteristic  curve  which  is  due  to  the  greater  margin 
between  the  temperature  at  which  the  gases  of  the  arc  conduct 
and  the  temperature  at  which  they  do  not,  to  the  greater  volume 
of  volatile  material  in  the  arc  and  the  lower  vaporization  point 
of  this  material. 

The  flame  carbon  arc  is  inherently  a  large  energy  light  unit. 
Since  the  light  is  emitted  quite  largely  by  the  flame,  it  is  desirable 
to  have  the  flame  as  long  as  possible.  Since  the  drop  at  the 
electrodes  is  fixed  at  approximately  30  volts  and  since  the  resist- 
ance drop  of  the  flame  is  approximately  38  volts  per  inch  (for 
a  10  ampere  arc)   it  will  be  evident  that  a  comparatively  high 


Fig.  10.— Oscillograms  of  a  flame  carbon  arc. 


voltage  is  desirable  for  most  economical  operation.  This  state- 
ment, together  with  all  other  statements  regarding  the  operating 
characteristics  of  the  arc,  should  be  considered  as  applying  to 
the  commercial  type  of  carbons  at  present  on  the  market,  and  it 
should  be  kept  in  mind  that  it  is  possible  to  materially  vary  these 
operating  characteristics  by  employing  special  types  of  carbons. 

The  present  current  densities  which  have  been   found  most 
satisfactory  are  as  follows  : 

Per  square  inch. 
100  amps. 

{1)     Current  density  at  the  arc    100 

(2)  Current  density  at    the  arc,   negative  terminal,   ap- 

proximately   1 ,500 

(3)  Current  density  at  the  arc,  positive  terminal,   with 

diffused  arc - 35 

\\  ith  commercial  carbons,  best  results  are  secured  with  approx- 


l68  TRANSACTIONS  I.  E.  S. — PART  II 

imately  10  amperes  through  the  arc  and  it  is  upon  this  basis 
that  the  above  statements  are  made.  Since  the  power  factor  of 
the  flame  carbon  arc  is  approximately  85  per  cent.,  it  will  be 
evident  that  the  wattage  under  which  most  satisfactory  operation 
may  be  obtained  lies  between  350  and  500  watts,  more  economical 
operation  being  obtained  with  a  higher  wattage. 

CONSIDERATION  OF  FLAME  CARBONS. 

As  the  carbons  mark  one  of  the  greatest  differences  between  the 
present  flame  carbon  arc  lamps  and  the  old  solid  carbon  arc 
lamps,  some  space  may  properly  be  devoted  to  their  consideration. 

The  flame  carbon  differs  from  the  solid  carbon  in  that  certain 
chemicals  are  added  during  their  manufacture  with  the  object 
of  increasing  the  light;  securing  better  operation,  and  reducing 
slag  troubles.  The  chemicals  commonly  added  may  be  classified 
as  follows : 

( 1 )  Illuminants. — These  comprise  the  compounds  of  three  or 
four  peculiar  elements.  For  white  light,  cerium  and  titanium 
offer  the  greatest  possibilities ;  while  calcium  and,  to  some  extent, 
tungsten  form  the  main  illuminants  in  yellow  light  carbons. 

(2)  Sustainers. — Since  the  flame  carbon  arc,  as  illustrated  in 
Figs.  2,  3  and  4  is  very  considerably  longer  than  the  enclosed 
carbon  arc,  it  would  be  prohibitively  unstable  unless  certain  com- 
pounds were  added  to  remedy  this  defect.  At  present,  fluorides 
and  to  a  small  extent,  borates,  form  the  most  suitable  sustainers 
when  slag  troubles,  evaporation  point  and  cost  are  considered. 
These  elements  are  frequently  introduced  as  compounds  of  the 
illuminants. 

(3)  Conductors. — In  order  to  overcome  certain  troubles  which 
arise  due  to  slag  which  may  form  in  case  the  carbon  is  consumed 
more  rapidly  than  the  illuminants  and  sustainers,  certain  com- 
pounds are  frequently  added  to  prevent  slag  troubles  which 
might  otherwise  result. 

By  a  series  of  long  experiments — for  theory  does  not  seem  to 
be  at  present  sufficiently  developed  to  do  much  more  than  indi- 
cate the  direction  of  progress — the  large  carbon  manufacturers 
have  to-day  developed  flame  carbons  which  are  entirely  commer- 
cial and  which  give  excellent  results,  both  from  the  standpoint 
of  efficiency  and  operation. 


darrah:    the  flame  carbon  arc  lamp  169 

In  view  of  the  scarcity  of  consistent  data,  it  is  perhaps,  dan- 
gerous to  venture  a  theory  to  account  for  the  high  efficiency  of 
the  flame  carbon  arc.  One  explanation,  however,  which  seems 
reasonable  and  which  accounts  for  a  large  number  of  facts  which 
have  been  observed,  is  to  consider  the  arc  as  a  place  where  the 
mineral  compounds  with  which  the  carbons  are  impregnated  are 
raised  to  such  a  temperature  that  dissociation  is  continually  in 
progress  and  that  at  very  short  distances  from  these  points  of 
high  temperature,  re-combination  is  occurring  very  vigorously. 
In  this  process  of  dissociation  and  re-combination,  the  elements 
which  form  the  illuminants  become  raised  to  the  temperature 
necessary  for  selective  radiation. 

A  consideration  of  the  position  in  Mendeleff's  table  of  the 
elements  which  form  the  illuminants  of  the  arc.  and  of  the 
chemical  characteristics  of  these  elements,  makes  the  above  theory 
appear  more  probable.  In  addition  to  the  high  temperature 
necessary  to  allow  selective  radiation,  it  is  essential  that  the 
illuminant  be  composed  of  a  material  which  readily  forms  com- 
pounds having  a  very  high  melting  point  and  vaporization  point. 
All  of  these  requirements  seem  to  be  met  by  the  oxides  of  cerium, 
calcium  and  tungsten.  In  this  connection — as  previously  pointed 
out — the  effect  of  cooling  the  arc  and  thereby  changing  its  color 
from  white  to  yellow  is  very  suggestive. 

Fig.  10a  shows  the  variation  of  candle-power  which  occurs 
with  increased  voltage  on  a  10-ampere.  alternating  current  flame 
carbon  arc  between  various  varieties  of  carbons.  These  curves 
are  interesting  as  indicating  that,  with  few  exceptions,  the  candle- 
power  increases  directly  with  the  voltage  until  a  certain  value  is 
reached,  and  in  some  cases  somewhat  more  rapidly.  This  is  to 
be  expected,  as  the  increased  voltage  allows  an  increased  arc 
length  and  therefore  a  longer  flame.  When  a  certain  value  of 
voltage  and  a  certain  arc  length  has  been  exceeded,  additional  cool- 
ing effects  introduced  prevent  a  material  increase  in  the  total  light 
flux  emitted,  and  for  this  reason,  the  curve  flattens  out  after  pass- 
ing a  certain  critical  point. 

Fig.  1 1  shows  the  effect  of  increased  currents  on  a  48-volt, 
60-cycle  arc  under  normal  conditions.  It  will  be  noted  that  the  light 
emitted  increase  uniformly  with  the  current  over  the  majority  of 


170 


TRANSACTIONS  I.  B.  S. — PART  II 


the  range  given.  Depending  upon  the  amount  of  illuminants  in 
the  current  increases.  This  affords  a  rough  means  of  determining 
the  amount  of  illuminant  in  a  given  grade  of  carbons.  In  this 
connection,  it  is  interesting  to  note  that  the  addition  of  illuminants 
to  a  flame  carbon  does  not  increase  the  total  light  emitted  in  a 
direct  ratio  to  the  amount  of  illuminants  added.  In  other  words, 
as  the  percentage  of  cerium  oxid  in  a  white-light  carbon  is  in- 
creased, the  candle-power  first  increases  until  a  maximum  value 


Fig.  10a. — Variation  in  candle-power  with  increase  of  voltage  across  arc 
of  flame  carbons. 


is  reached,  after  which  the  addition  of  more  cerium  oxid  will 
decrease  the  intensity  of  the  light  emitted  for  a  given  amount  of 
electrical  energy.  This  "condition  together  with  the  effect  of  slag, 
which  may  be  very  serious  in  improperly  designed  carbons,  limits 
the  amount  of  mineral  material  which  may  be  added  to  the 
carbons. 

The  difficulties  introduced  by  slagging  of  the  flame  carbons  has 


DARRAH  :      THE  FLAME  CARBON  ARC  LAMP 


171 


received  considerable  attention  and  has  been  overcome  com- 
mercially. Assuming  that  these  carbons  are  properly  designed, 
the  outage  during  normal,  commercial  operating  conditions  should 
not  exceed  y2  to  $4  of  1  per  cent.,  provided  the  lamps  are  main- 
tained in  the  proper  condition. 

Slagging  may  be  caused  by  the  entrance  of  air  into  the  arc 
chamber,  due  to  imperfect  globe  seats  or  leaking  condensers.  The 
excess  of  air  causes  the  consumption  of  the  carbons  to  proceed 
more   rapidly  than   the  mineral   components   can   be   vaporized. 


Kig.  11. — Variation  of  candle-power  with  increase  of  current  of  arc 
with   flame  carbons. 


This  allows  an  accumulation  of  fused  oxids  and  fluorids  upon  the 
surfaces  of  the  electrodes,  thus  forming  when  cold,  an  insulating 
layer.  An  excessively  low  current  or  an  arc  longer  than  normal 
may  also  cause  slagging. 

None  of  the  difficulties  mentioned  above  should  be  encountered 
in  the  commercial  lamps  providing  proper  globes  are  used  and  the 
lamps  are  maintained  in  good  condition.  A  very  small  amount 
of  slag  is  no  detriment  to  the  operation  of  the  flame  carbon  lamp 


172  TRANSACTIONS  I.  E.  S. — PART  II 

since  the  hammer  blow  which  is  given  when  the  carbons  fall  to- 
gether prevents  trouble  of  this  nature  from  being  serious. 

From  a  mechanical  standpoint,  it  may  be  stated  that  the  car- 
bons now  commercially  obtainable  are  entirely  satisfactory. 
Present  diameters  vary  from  approximately  24  m-  to  ]4  in-> 
depending  upon  the  conditions  under  which  they  are  used.  It  is 
possible  to  secure  these  carbons  within  0.025  in.  from  the 
specified  diameter  without  adding  appreciably  to  their  cost.  Since 
this  is  true,  it  will  be  evident  that  a  ring  clutch  may  be  employed 
in  the  operation  of  flame  carbon  lamps  with  entire  satisfaction. 

Since  the  average  flame  carbon  contains  more  than  30  per  cent, 
of  solid  material  which  is  not  consumed,  but  which  must  be 
vaporized  by  the  arc,  the  problem  of  disposing  of  this  material 
naturally  required  solution.  It  was  found  that  this  solid  material 
(called  "soot"  by  lamp  operators)  would  not  condense  upon  sur- 
faces which  were  maintaind  at  an  elevated  temperature,  while  it 
will  readily  condense  in  the  form  of  a  soft  white  powder  upon 
any  cool  or  relatively  cool  surface.  This  is  taken  advantage  of 
by  so  arranging  the  globes  of  the  lamps  that  they  will  be  ma- 
terially hotter  than  a  second  and  communicating  chamber  into 
which  the  gases  from  the  arc  are  forced. 

The  second  chamber,  which  is  made  of  metal  and  designed  to 
expose  a  comparatively  large  surface  to  the  air,  is  called  the  "con- 
densing chamber"  (Fig.  12),  shows  an  outline  sketch  of  a  flame 
carbon  lamp  and  indicates  the  path  of  the  gases  from  the  arc.  It 
will  be  noted  that  this  lamp  is  so  designed  that  the  majority  of 
the  gases  which  carry  the  vaporized  mineral  material  pass  directly, 
into  the  condensing  chamber  where  the  majority  of  the  soot  is 
deposited.  The  condensing  chamber  also  contains  sticks  of 
magnesia  or  other  alkaline  material  which  will  readily  combine 
with  the  hydrofluoric  acid,  nitric  acid,  etc.,  preventing  the  free 
acids  from  attacking  the  globe,  thereby  decreasing  its  transparency 
and  the  amount  of  light  emitted. 

An  examination  of  the  arc  through  an  absorbent  glass  shows 
a  very  rapid  movement  of  the  gases  immediately  surrounding  the 
arc.  Streams  of  semi-luminous  vapors  may  be  noted  being  pro- 
jected violently  upward  from  the  arc  flame.     These  vapors  con- 


DARRAH  :      THE  FLAME  CARBON  ARC  LAMP 


173 


tain  the  so-called  "soot,"  together  with  hydrofluoric,  nitric  and 
sulphuric  acids. 

Since  the  temperature  of  the  arc  in  its  cooler  portions  prob- 
ably exceeds  2.500  deg.  C,  while  the  remainder  of  the  arc 
chamber  is  filled  with  gas  at  a  temperature  not  exceeding  150 
deg.  C,  it  is  obvious  that  the  different  density  of  the  two  gases 
is  sufficient  to  account  for  the  violent  motion  in  the  vicinity  of 
the  arc. 

In  connection  with  the  wash  of  air  from  the  arc  around  the 
upper  carbon  is  found  an  explanation  of  the  markedly  different 


Fig.  12. — Sketch  indicating  path  of  gases  from  flame  carbon  arc  in  one  type  of  lamp. 

rates  of  consumption  of  the  upper  and  lower  carbons.  In  other 
words,  the  surface  of  the  upper  carbon  over  which  the  hot  gases 
pass  is  very  rapidly  consumed  by  any  free  oxygen  which  may  be 
in  the  arc  chamber. 

The  life  of  a  flame  carbon  depends  upon  its  density,  the  cur- 
rent, the  length  of  the  arc,  the  composition,  and  above  all,  the 
tightness  of  the  enclosure  in  which  the  -carbon  is  consumed. 
Every  effort  should  be  made  to  maintain  the  enclosure  as  perfect 
as  possible  at  all  times.  Only  accurately  ground  globes,  free 
from  flaws  at  the  globe  seat,  should  be  used ;  while  the  portion 
of  the  lamp  which  comes  in  contact  with  the  globe  should  be 
made  from  a  hard,  strong  alloy  which  will  be  affected  as  little 


174 


TRANSACTIONS  I.  E.  S. — PART  II 


as  possible  by  the  fumes  from  the  arc.  To  insure  maximum  life, 
it  is  necessary  that  all  joints  in  the  condensing  chamber  be 
maintained  tight  and  that  the  condenser  be  removed  as  little  as 
possible.  Under  commercial  conditions,  a  life  of  from  15  to 
20  hours  per  inch  may  be  obtained  from  standard  carbons. 

LAMP  PERFORMANCE. 

Passing  from  the  somewhat  more  theoretical  considerations  of 
the  arc  to  the  more  practical  point  of  operation  and  performance, 


Distribution  Curve  of  Clutch  Type  Fume- Carbon  Arc  Lamps 

Clear  Inner  and  A/bo  Outer  Globes.  White  Carbons. 


Terminal  Volts     £2 
Terminal  Watts  425 
/Ire  Volts . ...     46 
Arc  Watts 395 

Terminal  Volts       J/S 
Terminal  Walts     747.  S 

Arc  Volts 70 

Arc  lA/atls 455 


A.  C.  S£f?/£S 

Amperes 10 

EfT/c/ency...:..93X 
Power  Factor     816% 

D.C.  Multiple 

Amperes 6.5 

Efficiency.. ..6l/i 


M.L.H.C.P.....  .1000 

M.S.C.P. 656 

M.LUCP Permit  235 
M.5.CP   ■■    ■■    1.54 

M.l.H.CP. 820 

M.S.CP. 500 

M.LH.C  P  Per  Wolt  1.09 
MS.CP.     ■     "6? 


Fig.  13.— Comparative  distribution  curves  of  a  direct  current,  6.5  ampere  multiple 
and  a  10-ampere  series  alternating  current  flame  carbon  arc  lamp. 


the  distribution  curve  of  the  flame  arc  deserves  some  consider- 
ation. 

Fig.  13  shows  comparative  distribution  curves  of  the  direct 
current,  6.5-ampere,  multiple  lamp  and  the  alternating  current, 
10-ampere,  series  lamp.  It  will  be  noted  that  the  two  distribu- 
tion curves  are  very  similar.  The  slight  difference  in  shape 
between  the  distribution  curve  of  the  alternating  current  and 
the  direct  current  lamps  is  due,  mainly,  to  the  long  arc  of  the 


DARRAH  :      THE  FLAME  CARBON  ARC  LAMP  175 

direct  current  lamp  and  to  the  somewhat  greater  intrinsic  bril- 
liancy of  the  lower  portion  of  the  direct  current  arc. 

It  will  be  noted  that  these  distribution  curves  indicate  that  a 
very  considerable  amount  of  light  is  radiated  at  angles  above 
45  deg.,  the  maximum  illumination  occurring  at  approximately 
30  deg.  below  the  horizontal.  For  street  lighting  or  for  the 
illumination  of  large  areas,  this  is  a  very  desirable  characteristic, 
and  in  comparing  various  lamps  of  the  same  general  types,  it  is 
often  desirable  to  rate  the  lamps  on  such  a  basis  that  the  candle- 
power  between  the  horizontal  and  30  deg.  below  the  horizontal 
will  be  given  more  weight  than  the  candle-power  at  other  angles. 
In  other  words,  the  total  useful  light  flux  represented  by  the 
candle-power  at  any  given  angle  varies  with  the  cube  of  the 
cosine  of  the  angle. 

From  an  esthetic  standpoint,  the  flame  arc  is  particularly  desir- 
able because  of  the  relatively  low  intrinsic  brilliancy  of  the  light 
source  and  the  comparatively  large  area  from  which  the  light  is 
emitted ;  and  this  makes  an  illumination  from  the  arc  softer  and 
less  fatiguing  to  the  eye.  It  also  avoids  to  a  large  extent  the 
harsh  shadows  which  are  so  prominent  with  the  enclosed  carbon 
arc.  In  spite  of  the  lower  intrinsic  brilliancy  of  the  flame  carbon 
arc,  it  is  desirable  to  employ  the  diffusing  glassware. 

The  distribution  of  energy  in  the  spectrum  of  the  flame  carbon 
arc  is  peculiarly  well  suited  for  illumination  which  is  intended  to 
approximate  daylight,  although  where  the  matching  of  color 
values  is  of  special  importance  the  carbons  employed  should  be 
designed  with  this  point  in  view.  In  the  light  from  white  car- 
Dons,  the  red  and  green  wave  lengths  are  found  in  very  materi- 
ally greater  proportions  than  in  the  enclosed  carbon  arc  or 
metallic  flame  arcs. 

The  efficiency  of  the  flame  arc  as  an  illuminant  is  particularly 
high.  The  mean  lower  hemispherical  candle-power  per  watt 
varies  from  2.5  with  white  carbons  and  "Alba"  glassware,  to 
5  mean  lower  hemispherical  candle-power  per  watt  with  clear 
glassware.  When  opalescent  globes  are  employed,  the  efficiency 
falls  between  these  values.  Yellow  carbons  may  be  obtained 
which  give  from  30  to  50  per  cent,  more  light  than  the  white 
carbons. 
3 


I76  TRANSACTIONS  I.  E.  S. — PART  II 

From  an  operating  and  maintenance  standpoint,  the  flame 
carbon  lamp  is  peculiarly  suited  for  the  illumination  of  city 
streets,  as  well  as  for  factory  lighting.  The  lamps  require  but 
little  attention ;  the  life  between  trims  is  approximately  100 
hours,  and  the  time  and  labor  of  trimming  the  lamps  is  small. 
Under  efficient  direction,  one  man  could  very  readily  trim  and 
clean  from  50  to  75  lamps  per  day.  At  each  trimming  period,, 
it  is  desirable  to  carefully  remove  the  "soot"  from  the  globes 
and  condensing  chamber,  while  after  each  1,000  hours  of  opera- 
tion, the  consumer  should  be  renewed.  Depending  upon  the 
care  and  efficiency  with  which  the  installation  is  maintained,  the 
life  of  the  globes  varies  from  2,000  to  5,000  hours.  The  impor- 
tance of  maintaining  the  globes  clean  and  in  good  condition  is 
frequently  not  sufficiently  appreciated  by  operating  engineers,, 
and  as  a  result,  the  light  efficiency  of  the  lamp  is  very  materially 
reduced. 

CONCLUSIONS. 

A  consideration  of  the  points  noted  above  will  make  evident 
some  of  the  features  which  have  been  incorporated  in  modern 
flame  carbon  arc  lamps  and  some  of  the  difficulties  which  have 
been  encountered  and  commercially  overcome.  At  the  present 
time,  the  flame  carbon  arc  lamp  is  a  commercial  success  and  has 
stood  the  test  of  actual  operating  conditions.  The  field  of  the 
flame  carbon  arc  lamps  is  continually  expanding,  and  due  to 
the  increased  prosperity  of  the  period  and  the  fact  that  engi- 
neering development  has  kept  pace  with  commercial  require- 
ments, there  seems  to  be  little  doubt  but  that  the  installation  of 
flame  carbon  lamps  will  in  the  near  future  considerably  out- 
distance the  installation  of  all  other  types  of  large  energy  light- 
ing units,  including  those  used  in  the  most  utilitarian  and  busiest 
shops  and  factories,  as  well  as  those  employed  for  highly  orna- 
mental street  or  parkway  lighting. 


THE  FLAME  CARBON  ARC  LAMP  1/7 

DISCUSSION. 

Messrs.  W.  R.  Mott,  R.  B.  Chillas  Jr.,  and  A.  T.  Baldwin 
(communicated)  :  Several  points  in  Mr.  Darrah's  excellent 
paper  bring  up  questions  which  we  desire  to  discuss  briefly. 

Along  with  the  other  differences  mentioned  between  the  flame 
arc  and  the  magnetite  arc  it  may  be  well  to  bring  attention  again 
to  the  fact  that  the  nature  of  the  positive  flame  electrode  largely 
determines  the  color  and  candle-power  of  the  arc,  while  with  the 
magnetite  arc  the  negative  is  the  determining  factor. 

An  examination  of  oscillograms  of  flame  arcs  will  show  that 
the  less-than-unity  power  factor  is  the  result  of  a  distortion  rather 
than  a  displacement  of  the  waves.  The  high  power-factor  of  the 
flame  arc  is  due  to  the  ease  with  which  the  current  is  re-established 
through  the  hot  gases  after  each  reversal  of  current.  Since  the 
light  flux  is  largely  dependent  upon  the  current  it  is  evident  that 
the  least  variation  or  flicker  in  the  light  will  be  found  in  the  arc 
of  highest  power  factor.  As  indicated  in  the  paper  the  starting 
voltage  of  the  flame  arc  is  very  low.  In  fact  it  is  much  lower  for 
the  flame  arc  than  for  either  the  enclosed  arc  or  magnetite.  In 
these  respects  the  flame  arc  in  general  has  a  marked  advantage 
over  the  other  arc  illuminants.  Certain  types  of  commercial  car- 
bons possess  these  properties  to  a  far  greater  degree  than  others. 
Such  carbons  are  particularly  useful  on  low  frequency  (25  cycle) 
circuits,  and  in  lighting  shops  containing  moving  machinery, 
especially  if  the  movement  of  machine  parts  and  the  cyclic  varia- 
tion of  the  light  are  nearly  synchronous. 

The  alternating  current  flame  arc  possesses  one  rather  striking 
advantage  over  the  plain  carbon  arc  because  it  shows  less  flicker 
at  lower  frequencies.  In  the  plain  carbon  arc  the  electrode  which 
is  positive  at  any  instant  is  giving  out  most  of  the  light.  This 
means  that  on  25  cycles  the  light  in  one  direction  passes  through 
a  maximum  25  times  per  second.  In  the  flame  arc,  where  the 
light  comes  from  the  arc  stream  itself,  the  corresponding  figures 
are  50  times  per  second,  or  far  beyond  the  critical  flicker 
frequency  of  the  eye. 

The  best  arc  voltage  to  be  used  is  determined  to  a  large  extent 
by  other  factors  taken  in  conjunction  with  those  pointed  out  by 
Mr.  Darrah.     As  the  arc  voltage  and  consequently  the  arc  length 


I78  TRANSACTIONS  I.  £.  S. — PART  II 

is  increased  at  constant  current  a  point  is  reached  at  which  the 
candle-power  increase  is  not  as  fast  as  the  voltage  increase  due  to 
the  greater  cooling  of  the  arc  and  the  steadiness  begins  appreci- 
ably to  diminish.  Carbons  can  be  made  which  will  operate  satis- 
factorily at  the  higher  arc-voltages  but  such  operation  is  obtained 
at  the  sacrifice  of  candle-power  due  to  the  increased  amount  of 
"sustainers"  required  at  the  expense  of  the  quantity  of 
"illuminants"  present.  This  follows  from  the  fact  that  the  total 
amount  of  flaming  materials  in  the  carbons  is  limited  by  the 
slagging  tendency  so  that  an  increase  of  one  type  of  material 
must  be  accompanied  by  some  decrease  in  the  others.  Taking  all 
things  into  consideration,  it  will  be  found  that  the  highest  candle- 
power,  the  best  steadiness  and  reliability  will  be  obtained  in  no 
volt  multiple  lamps  when  the  arc  voltage  does  not  exceed  65  and 
the  current  not  lower  than  6  amperes.  For  series  lamps  and 
alternating  current  multiple  lamps  with  auto-transformers,  40  to 
45  arc  volts  gives  good  service  when  the  current  is  not  less  than 
10  amperes.  In  general,  however,  a  flame  carbon  can  be  made 
to  operate  satisfactorily  commercially  over  a  very  wide  range  of 
current  and  voltage  for  a  given  lamp,  but  at  present  there  is  no 
one  carbon  that  will  meet  all  conditions  satisfactorily  in  every 
way  in  all  lamps. 

The  current  densities  within  the  arc  will  be  found  to  vary  quite 
widely.  The  cross-section  which  the  arc  assumes  is  that  which 
will  maintain  the  equilibrium  between  the  energy  input  and  the 
output  of  light  and  heat.  The  size  depends  on  so  many  factors, 
such  as  the  mutual  attraction  of  the  current  filaments  in  the  arc 
stream,  the  volume  of  gas  produced  by  all  of  the  different 
chemicals  at  the  various  temperatures  and  pressures,  the  repulsion 
of  similarly  charged  electron,  that  we  can  say  only  that  the  given 
current  densities  have  been  found.  That  they  are  also  satisfactory 
is  very  fortunate.  The  diffused  arc  mentioned,  is  an  effect 
obtained  only  occasionally  in  some  direct  current  arcs ;  never  in 
alternating  current  arcs.  It  is  of  much  lower  efficiency  than  the 
normal  direct  current  arc  which  has  a  positive  crater  current- 
density  of  the  order  of  300  amperes  per  square  inch.  We  have 
found  the  usual  tendency  to  be  that  arcs  of  the  higher  current 
densities  eive  higher  efficiencies. 


THE  FLAME  CARBON  ARC  EAMP  1/9 

The  use  of  magnesia  blocks  in  the  condensing  chamber  to  pre- 
vent globe  etching  can  hardly  be  considered  successful.  Toward 
the  end  of  the  trim-life  the  blocks  become  covered  with  dust 
which  prevents  the  complete  absorption  of  the  harmful  gases  and 
so  permits  them  to  attack  the  globes.  The  blocks  must  be  re- 
newed from  time  to  time,  which  introduces  an  additional  cost  into 
the  maintenance  account.  With  properly  designed  carbons,  the 
magnesia  blocks  can  be  omitted  and  globe  etching  does  not  occur. 

Mr.  F.  A.  Vaughn  (communicated)  :  Just  one  thought 
appears  in  connection  with  the  author's  reference  in  his  first 
paragraph  to  the  relegation  to  the  scrap  pile  of  old  equipment, 
especially  enclosed  arc  lamps,  and  old  types  of  the  prismatic  glass 
reflector.  This  very  point  is  sometimes  a  matter  of  considerable 
consequence  to  illuminating  engineers  who  wish  to  do  their  duty 
not  only  to  their  client  in  any  particular  case  but  to  the  common- 
wealth at  large  as  far  as  the  great  movement  for  conservation  of 
vision  is  concerned. 

The  client  naturally  wishes  to  obtain  as  great  an  advantage  in 
the  disposal  of  the  old  equipment  as  possible  and  the  suggestion 
is  almost  always  presented  that  these  units  and  these  reflectors 
are  good  enough  for  other  installations,  and  he  usually  makes 
every  effort  possible  on  his  own  part  and  desires  the  engineer  to" 
do  so  also,  to  dispose  of  the  equipment  to  other  unsuspecting  and 
unknowing  users  of  illumination.  The  illuminating  engineer's 
duty  to  his  client  is,  perhaps,  to  obtain  the  advantage  for  his  client 
in  so  doing,  but  his  duty  to  the  commonwealth  is  to  use  a  strong, 
heavy  sledge  hammer  on  the  entire  equipment  and  thus  put  an  end 
to  its  career  as  a  destroyer  of  human  vision,  not  only  as  far  as 
that  specific  client  is  concerned  but  also  the  entire  universe. 

It  thus  devolves  upon  the  illuminating  engineer  to  convert  his 
client  to  enough  sympathy  in  the  conservation  of  vision  move- 
ment to  feel  willing  to  allow  the  whole  equipment  to  be  disposed 
of  for  all  time. 


l80  TRANSACTIONS  I.  E.  S. — PART  II 

THE  ILLUMINATION  OF  MOTION  PICTURE 
PROJECTORS.* 


J.   FRANK  MARTIN. 


Synopsis:  After  outlining  the  operation  characteristics  and  require- 
ments of  the  light  sources  of  projection  lamps,  the  color  of  light,  and  the 
character  of  screen  best  adapted  to  projecting  motion  pictures,  this  paper 
is  concluded  with  a  brief  discussion  of  the  question,  Does  the  motion 
picture  cause  eyestrain?  Observation,  the  author  states,  indicates  that 
constant  viewing  of  motion  pictures  tends  not  only  to  develop  the  semi- 
voluntary  muscles  of  the  eye  but  to  give  them  greater  endurance  and 
more  rapid  action.  He  adds,  though,  that  pronounced  flicker  in  pictures 
may  lead  to  over-stimulation  and  injury.  The  discomfort  and  irritation 
sometimes  experienced  by  patrons  of  motion  pictures,  caused  by  unsteadi- 
ness of  the  pictures  and  defects  in  the  films,  indicates  so  much  fatigue 
which  usually  recedes  without  any  resulting  injury.  The  author  suggests 
an  examination  of  the  eyes  of  lantern  operators  to  determine  the  nature 
and  extent  of  the  effect  of  motion  pictures  on  the  eye. 

The  motion  picture  is  now  recognized  as  an  educational  factor 
second  in  importance  only  to  the  printing  press.  It  is  a  develop- 
ment of  the  last  twelve  years  and  like  other  great  inventions 
which  possess  novelty  and  meet  a  demand  of  the  public,  the  in- 
troduction has  been  accompanied  by  a  disregard  of  scientific 
principles  and  the  necessary  standardization  to  secure  the  greatest 
efficiency. 

The  magnitude  of  the  industry  may  be  judged  by  a  considera- 
tion of  the  fact  that  there  are  more  than  twenty  thousand  motion 
picture  theatres  in  America,  having  an  average  daily  attendance 
of  twenty  million  people  and  a  maximum  demand  for  electric 
current  in  excess  of  sixty  thousand  kilowatts,  which  about  equals 
that  for  lighting  a  city  such  as  Philadelphia,  Boston  or  St.  Louis. 

This  attendance  of  these  theatres  compared  with  the  total  num- 
ber of  people  who  are  habitual  readers  by  means  of  artificial  light 
shows,  that  if  the  general  impression  that  the  motion  picture  seri- 
ously affects  the  eye  is  correct,  the  illuminating  engineer  is  con- 
fronted with  a  problem  which  is  not  receiving  the  attention  which 
its  importance  demands. 

*  A  paper  read  before  a  meeting  of  the  Pittsburgh  section  of  the  Illuminating  Engi- 
neering Society,  April  iS,  1913. 


martin:    illumination  of  motion  picture  projectors  181 

THE  SOURCE  OF  LIGHT  FOR  PROJECTION. 
From  the  beginning,  wherever  electric  current  was  available, 
the  electric  arc  has  been  the  only  source  of  light  considered.  The 
requirements  for  a  satisfactory  projector  iluminant  are:  first,  as 
near  an  approach  to  a  point  source  of  light  as  is  possible ;  and, 
second,  the  most  intense  source  of  light  available.  The  carbon 
arc  is  the  nearest  approach  to  these  requirements,  and  the  modern 
motion  picture  projector  has  practically  been  built  around  the 
electric  arc. 

CHARACTERISTICS  OF  THE  PROJECTOR  ARC. 
Fig.  i  shows  the  voltage  at  the  arc  on  both  direct  and  alternat- 
ing current,  using  a  type  of  carbon  which  is  widely  used,  at  vary- 
ing current  densities.     The  arc  gap,  size  of  carbons  and  align- 


40 


o30 


D.C. 

/ 

1 

t^S 

■-' 

FLAN 

E  A. 

c 

A.C. 

j 

5      10     15     20    25    30    35    40     45    50    5S 
AMPERES 

Fig.  i. — Characteristics  of  projector  arcs. 

ment  of  carbons  was  adjusted  for  each  current  value  so  as  to  give 
the  most  stable  performance.  The  dotted  line  in  Fig.  i  repre- 
sents a  special  flaming  arc  carbon,  and  the  curve  is  representative 
for  both  alternating  and  direct  current. 

ALIGNMENT  AND  DIMENSIONS  OF  CARBONS. 
Fig.  2  illustrates  two  methods  of  aligning  the  carbons  and  the 
distribution  of  light  resulting  therefrom.  These  settings  repre- 
sent the  extremes  between  which  satisfactory  performance  of  the 
arc  can  be  secured.  On  direct  current  the  top  carbon  must  in- 
variably be  made  the  positive  electrode  of  the  arc  so  as  to  direct 


182 


TRANSACTIONS  I.  E.  S. — PART  II 


the  maximum  flux  of  light  in  a  horizontal  plane.  On  both  direct 
and  alternating  current,  the  top  carbon  gives  the  most  light  by 
virtue  of  its  position  in  the  draught  of  intensely  heated  glass 
from  the  arc.  The  top  carbon  also  burns  away  faster  from  the 
same  cause. 


'INCLINED  RIGHT  AN6LE 

Fig.  2.— Alignment  of  electrodes. 

It  is  very  important  that  the  size  of  the  carbon  and  the  relative 
diameter  of  the  core  and  shell  be  closely  regulated  to  the  current 
density.  Both  the  stability  of  the  arc  and  the  intensity  of  light 
may  be  materially  increased  by  varying  the  dimensions  of  the 
carbons  used. 

THE  MECHANISM  OF  PROJECTOR  LAMPS. 

The  modern  projector  lamp  retains  the  elementary  construction 
and  principle  of  the  most  primative  electric  light.  Fig.  3  and  4 
illustrate  the  most  approved  forms  in  use  at  the  present  time.  Hand 
operated  lamps  are  universally  used  on  account  of  it  being  im- 
possible to  construct  a  lamp  which  will  automatically  center  the 
rrc  in  the  lens  axis  and  compensate  for  the  wandering  of  the 
arc. 

THE  PERFORMANCE  OF  PROJECTOR  ARCS. 

Fig.  5  shows  the  relative  candle-power  of  the  arc  on  alternating 
and  direct  currents  at  varying  current  densities.  Readings  were 
made  in  a  horizontal  plane,  the  size  of  carbons,  alignment  and  arc 
gap  were  adjusted  in  each  case  to  give  the  steadiest  performance. 
Attention  is  called  to  the  termination  of  the  curves  for  the  right 
angle  and  flaming  arc  at  about  27  amperes,  beyond  which  point 
the  magnetic  blow  out  effect  in  this  method  of  aligning  the  elec- 


MARTIN  :     ILLUMINATION  OP  MOTION  PICTURE  PROJECTORS    183 

trodes  has  such  a  value  that  the  arc  is  lengthened,  becomes  very 
unsteady  and  consequently  unsatisfactory. 

For  the  same  current  density,  all  forms  of  the  arc  have  a  much 
greater  intensity  on  direct  current  than  on  alternating  current. 


Fig.  3.— Straight  line  lamp. 


Fig.  4. — Angle  lamp. 


This  condition  is  due  to  the  pronounced  crater  of  the  direct  cur- 
rent arc  and  the  ease  with  which  the  flux  of  light  from  this  crater 
may  be  directed  by  the  alignment  of  the  electrodes. 


1 84 


TRANSACTIONS  I,  E.  S. — PART  II 


COLOR  OF  LIGHT. 
The  color  value  of  the  light  used  in  projection  is  of  small  im- 
portance unless  it  is  considered  from  an  esthetical  viewpoint  in 
conjunction  with  photographic  processes.     However,  under  pres- 
ent conditions  the  combination  of  effects  arising  from  the  use  of 


3000 


2500 


£2000 


£  1500 


1000 


/ 

*i 

/ 

§ 

.* 

M 

4 

¥ 

4* 

?r 

0f 

W" 

<W 

u»°J? 

& 

i 

LAM 

M0 

" 

1 

2000        4000        6000         8000         10,000 
EFFECTIVE  CANDLE-POWER 

Fig.  5.— Performance  of  projector  arc. 


arcs  giving  a  bluish  light  and  very  contrasty  photography  gives 
exceedingly  harsh  impressions.  The  use  of  an  arc  in  which  yel- 
low is  the  predominating  color  would  make  a  material  improve- 
ment. 

THE  LENS  SYSTEM  OF  PROJECTORS. 
Many  different  combinations  of  lenses  have  been  experimentally 


FILM 


08"  PROJECTION 
[j=   .LENS 

m 


.-Hi 


SCREEN-) 


INTRINSIC 
BRILLIANCY 


200  OP' 


'-GO  OP    5I0C-P-"       :470C-P 


0.0022  C-P- 
0.4  FT.  -CANDLES 


LOSSES  IN  PER  CENT 


70 


20 


Fig.  6. — The  lens  system  of  projectors. 


developed  but  no  radical  changes  have  been  made  in  the  earliest 
form  of  lens  used  in  the  magic  lantern.     The  lens  system  and  the 


martin:    illumination  of  motion  picture  projectors  185 

losses  therein  are  illustrated  in  Fig.  6.  It  has  been  built  up  with 
a  point  sources  of  light  as  a  basis  consequently  the  low  efficiency 
of  10  per  cent,  is  not  surprising  and  there  is  apparently  a  great 
opportunity  for  improvement. 

REGULATION  OF  THE  PROJECTOR  ARC  AND  METHODS  OF 
CONVERTING  ALTERNATING  TO  DIRECT  CURRENT. 

Within  certain  limits  of  projection  which  may  be  described  as 
the  illumination  necessary  to  project  an  image  not  exceeding  one 
hundred  square  feet  in  area,  direct  current  through  a  resistance 
is  not  highly  preferable  to  alternating  current  supplied  through  an 
auto-transformer,  the  service  voltage  being  equal  in  each  case  at 

5000 


4000 


z3000 

z 
02000 


woo 


AjC.-IK)V.& 

:.mo" 

OfH 

ENER 

1T0R 

5ET- 

y 

/ 

/ 
/ 

A.CJ 

UTOT 

<mu 

"ORM 

*} 

/ 

m 

s 

~ 

/     , 

4 

■-P 

/ 

s 

*s 

■ 

-/ 

*L. 

SPECIAL  15  KW.  MOT 
GENERATOR  SET- 

)R- 

~^~ 

'* 

2000        1000        6000        8000 
EFFECTIVE  CANDLE-POWER 


10,000 


Fig.  7.— Performance  of  projector  arc  on  alternating  to  direct  curient  converters. 


1 10  volts.  Many  forms  of  self-regulating  motor  generator  sets, 
autotransformers  and  rectifiers,  all  having  a  regulation  char- 
acteristic closely  approaching  that  of  a  metallic  resistance,  are  in 
use.  Fig.  7  shows  the  comparative  efficiency  of  the  arc  operating 
in  conjunction  with  these  machines.  All  the  readings  were  taken 
after  setting  the  lamp  for  the  most  satisfactory  operation. 

The  majority  of  motion  picture  theatres  use  auto-transformers 
of  high  reactance,  which  are  manufactured  and  marketed  under 
various  trade  names. 

For  the  reason  that  the  use  of  alternating  current  gives  a  very 
low  efficiency  at  the  arc,  requiring  heavier  current  for  a  given 
illumination  and  an  accompanying  increase  in  heat  dissipated  and 
under  average  conditions  will  increase  the  inherent  flicker  of  the 


1 86  TRANSACTIONS  I.  £.  S. — PART  II 

projected  image,  it  is  highly  desirable  to  standardize  the  apparatus 
so  that  direct  current  is  supplied  to  the  projection  lamp.  A  care- 
ful analysis  indicates  that  this  apparatus  should  be  a  motor-gen- 
erator set,  the  motor  allowing  of  substitution  for  either  alter- 
nating or  direct  current  at  different  voltages,  and  the  generator 
having  a  capacity  of  3.5  kilowatts,  65  volts,  direct  current,  com- 
pound wound  and  designed  to  deliver  175  per  cent,  full  load  for 
a  five  minute  period  of  each  hour.  One  or  more  projection 
lamps,  each  connected  through  a  small  resistance  to  give  the 
necessary  ballast,  could  be  supplied  from  this  machine  and  could 
be  operated  simultaneously  for  short  periods  without  affecting 
the  other.  The  field  rheostat  and  lamp  ballast  should  be  placed 
conveniently  near  the  operator  so  that  the  intensity  of  light  can 
be  regulated  for  the  varying  density  of  the  films.  This  machine 
could  be  universally  used  with  an  improvement  in  economy  and 
quality  of  projection.  The  relative  efficiency  of  this  machine  is 
also  illustrated  in  Fig.  7. 

PROJECTION  SCREENS. 

The  low  intensity  of  illumination,  which  is  less  than  0.4  foot- 
candles  on  the  average  motion  picture  screen,  immediately  leads 
to  the  conclusion  that  improving  the  illumination  is  largely  a 
matter  of  increasing  the  intensity  of  the  light  source  and  the 
efficiency  of  the  lens  system.  This  conclusion  is  also  apparently 
substantiated  by  the  fact  that  the  difference  in  the  efficiencies 
of  the  available  reflecting  surfaces  allows  of  no  great  in- 
crease in  reflecting  efficiency.  However,  due  to  the  fact  that 
the  eye  is  very  much  more  susceptible  to  changes  in  illumina- 
tion at  low  intensities  than  at  those  ordinarily  experienced,  it 
is  possible  to  apparently  double  the  screen  illumination  by  in- 
creasing the  efficiency  of  the  reflecting  surface  only  10  per  cent. 

In  an  effort  to  satisfy  a  public  demand  for  more  wholesome 
and  better  lighted  surroundings,  there  has  been  developed  a 
number  of  improved  screens.  There  are  two  distinct  types  of 
these  screens :  one  consists  of  a  smooth  surface  dressed  with 
aluminum  powder  and  lacquer;  the  other  is  a  dressing  of  the 
whitest  and  most  opaque  paint  obtainable,  applied  to  a  smooth 
surface  or  to  the  back  of  a  large  piece  of  plate  glass.     The 


martin:    illumination  of  motion  picture  projectors  187 

aluminum  dressed  screen  closely  approaches  the  efficiency  of  a 
low  grade  amalgam  hacked  mirror  and  unless  the  surface  is 
pebbled  or  roughened,  the  image  on  the  screen  is  not  entirely 
discernible  to  a  spectator  sitting  to  one  side,  without  the  angle 
of  incidence.  This  screen  must  also  be  mounted  in  a  perfect  flat 
plane  to  avoid  a  serious  glare  and  on  account  of  the  metallic  tint 
gives  color  distortion.  A  piece  of  plate  glass,  backed  with  an 
opaque  white  surface,  although  costly,  is  the  ideal  screen.  This 
construction  gives  a  perfectly  smooth  surface,  fine  grained  and 
gives  the  highest  efficiency  allowable  in  securing  satisfactory 
definition.  White  oil  cloth  stretched  on  a  rigid  frame  is  less 
durable  but  almost  as  satisfactory. 

THE  ILLUMINATION  OF  AUDITORIUMS  WHERE  PROJECTION 

IS  USED. 

The  first  consideration  is  to  avoid  all  sources  of  light  which 
will  cause  glare.  Wall  brackets  fall  in  this  category  and  should 
never  be  placed  in  the  range  of  vision,  between  the  spectator  and 
the  screen.  Indirect  illumination  from  chandeliers  or  cove  light- 
ing is  highly  preferable  as  it  best  avoids  glare. 

The  intensity  of  illumination  should  be  graduated  from  the 
screen  outward  toward  the  rear  of  the  auditorium  and  should 
average  about  0.2  foot-candle  at  the  level  of  the  backs  of  seats. 
The  value  at  the  rear  where  the  movement  of  spectators  is  con- 
gested may  generally  be  increased  to  one  foot-candle  with  satis- 
factory results. 

POSSIBLE  METHODS  OF  IMPROVING  PROJECTION. 

Experiments  with  a  view  to  adapting  the  flaming  arc  to  pro- 
jection have  been  made  and  have  demonstrated  that  an  increase 
of  approximately  60  per  cent,  can  be  made  in  the  illumination  up 
to  a  current  density  of  20  amperes.  The  excessive  length  and 
large  area  of  the  flame,  which  give  practically  all  the  light,  is  a 
recession  from  the  desired  point  source  and  makes  it  necessary 
to  use  the  right  angle  alignment  of  electrodes  which  determines 
limitation  of  the  above  current  value. 

Another  method  allowing  of  an  increase  in  the  efficiency  of 
the  lens  system  appears  to  be  in  the  use  of  a  metallic  filament 
lamps  having  a  short  heavy  filament  arranged  with  a  reflector  and 


1 88 


TRANSACTIONS  I.  E.  S. — PART  II 


lens  as  shown  in  Fig.  8.  The  lamp  may  be  worked  at  a  consid- 
erable over-voltage  and,  by  interrupting  the  current  periodically 
in  synchronism  with  the  movement  of  the  film,  the  overshooting 


Fig.  8.— Special  lighting  system  for  projectors. 

phenomena  of  the  lamp  will  increase  the  illumination  and  the 
present  sector  shutter  can  be  eliminated. 

DOES  THE  MOTION  PICTURE  CAUSE  EYESTRAIN. 

It  is  believed  that  standardization  to  secure  commercial  effi- 
ciency should  be  considered  secondary  to  the  correction  of 
physiological  defects.  The  illuminating  engineer  in  his  work  in 
the  field  of  motion  picture  projection  should  give  particular 
attention  to  the  possibility  of  evil  effects  accruing  to  the  eyes 
of  those  who  witness  picture  performances. 

Before  going  further,  it  must  be  understood  that  the  follow- 
ing analysis  is  not  made  from  the  viewpoint  of  the  ophthalmolo- 
gist with  a  full  understanding  of  the  principles  of  physiology 
involved,  but  is  rather  from  the  viewpoint  of  the  layman.  If 
this  paper  serves  the  purpose  of  attracting  attention  to  the  sub- 
ject, it  will  have  served  its  mission. 

Eyestrain,  discomfort  and  permanent  injury  may  for  practical 
purposes  be  divided  into  two  classes :  causes  which  affect  the 
muscles  of  the  eye,  and  those  which  cause  deterioration  of  the 
nerve  structure. 


martin:    illumination  of  motion  picture  projectors  189 

In  considering  the  effect  of  motion  pictures,  two  sources  of 
evil  are  found  to  be  practically  absent.  First,  there  is  little  danger 
from  intense  light  entering  the  eye  and  there  is  no  glare  unless 
foreign  sources  of  light  are  within  the  range  of  vision ;  therefore 
there  can  be  no  deterioration  of  the  nerve  structure.  Second, 
when  viewing  the  image  on  the  screen  the  involuntary  muscles 
controlling  the  focus  of  the  eye  lens  are  inactive  and  there  can 
be  no  evil  influence  on  these  muscles.  This  leaves  two  possible 
conditions  which  may  have  an  injurious  effect.  The  most  com- 
mon fault  is  the  flicker  which  involves  the  basic  principle  of 
motion  picture  projection.  The  other  fault  lies  in  the  common 
but  unnecessary  evil  of  unsteadiness  in  the  image  on  the  screen, 
which  is  the  result  of  defective  methods  in  the  manufacture  of 
film  or  in  the  maintenance  of  the  projector  mechanism. 

The  effect  of  flicker  is  confined  to  the  involuntary  muscles 
which  control  the  action  of  the  iris.  The  natural  action  of  the 
iris  is  to  close  very  rapidly  and  open  slowly.  When  looking  at  a 
motion  picture  the  pupil  is  contracted  to  a  somewhat  smaller 
opening  than  when  viewing  a  still  picture  having  the  same 
illumination.  When  flicker  is  pronounced,  the  pupil  may  be 
observed  to  tremble  slightly  as  if  attempting  to  follow  the  fluctua- 
tion of  light.  This  may  lead  to  over-stimulation  and  permanent 
injury. 

The  discomfort  and  irritation  of  the  eye  sometimes  experienced 
by  the  patrons  of  motion  picture  theatres  is  due  to  the  un- 
steadiness of  the  picture  and  blemishes  of  the  film  used.  These 
faults  affect  the  semi-voluntary  muscles  which  control  the  move- 
ment of  the  eye  ball  and  the  discomfort  is  due  to  fatigue  which 
recedes  without  apparent  injury.  Observation  indicates  that  con- 
stant viewing  of  motion  pictures  tends  to  develop  these  muscles, 
giving  them  greater  endurance  and  more  rapid  action.  To  secure 
light  on  the  subject  it  is  suggested  that  observations  at  first  be 
confined  to  motion  picture  operators  who  as  a  class  of  trades- 
men do  not  show  the  evil  effects  that  would  be  expected. 


I90  TRANSACTIONS  I.  £.  S. — PART  II 

DISCUSSION. 

Dr.  Eluce  M.  Alger  (communicated)  :  In  discussing  Mr. 
Martins  very  interesting  paper  I  must  confess  at  the  outset  that 
my  personal  experience  with  moving  pictures  has  been  more 
theoretical  than  practical  because  to  me,  as  to  many  other  people, 
they  have  been  so  productive  of  strain  and  discomfort  and  head- 
ache that  I  avoid  them. 

It  does  not  of  course  necessarily  follow  that,  because  they  cause 
discomfort  and  fatigue,  that  they  also  cause  organic  damage  to 
the  eyes  provided  the  fatigue  is  kept  within  the  individuals 
physiological  limits.  That  point  can  only  be  determined,  as  Mr. 
Martin  suggests,  by  a  careful  observation  of  the  effects  on  the  eyes 
of  employees  and  others  whose  exposure  is  a  continuous  instead 
of  an  occasional  one. 

Neither  do  I  doubt  that  even  in  matters  of  fatigue  and  strain 
most  individuals  would  with  practise  develop  a  great  increase  in 
their  ability  to  compensate  for  them.  But  I  do  feel  reasonably 
sure  that  the  strain  imposed  by  frequent  watching  moving 
pictures,  when  superimposed  on  the  fatigue  caused  most  of  us 
by  our  customary  ocular  tasks,  very  frequently  passes  beyond 
physiological  limits  and  must  eventually  produce  deterioration. 

Furthermore  it  seems  to  me  that  many  of  the  factors  which 
cause  fatigue  cannot  be  done  away  with  but  are  inherent  in  the 
pictures  by  reason  of  their  being  "motion"  pictures. 

I  am  inclined  to  disagree  with  the  author's  opinion  that  there 
can  be  no  strain  of  any  account  caused  by  the  action  of  the  ciliary 
muscles  in  focussing,  since  the  screen  is  some  distance  away  from 
the  eyes  and  that  distance  an  unchanging  one.  Even  with  a 
perfectly  normal  eye  the  intentness  with  which  one  must  watch 
a  moving  picture  prevents  in  large  measure  the  complete  muscular 
relaxation  which  at  frequent  intervals  should  rest  our  eyes.  But 
very  few  of  us  have  normal  eyes.  Most  of  us  are  far  sighted  or 
astigmatic  and  many  of  us  have  muscular  defects  as  well.  We 
have  to  focus  to  see  even  distant  details  distinctly  and  while  this 
muscular  effort  need  not  be  very  great  for  the  moment  the  con- 
stant rapid  changes  on  the  screen  make  it  almost  continuous. 

In  real  life  we  see  things  momentarily  much  as  a  painter  would 
paint  them.     Things  in  the  plane  for  which  our  eyes  are  adjusted 


ILLUMINATION  OF   MOTION   PICTURE   PROJECTORS  I9I 

are  clear  and  distinct  while  those  out  of  focus  are  more  or  less 
blurred,  so  that  we  may  be  only  partly  conscious  of  them.  In  the 
moving  picture,  things,  which  in  nature  were  in  different  planes, 
are  reproduced  on  the  one  plane  of  the  screen  and  so  in  a  sense 
are  forced  upon  our  consciousness  simultaneously.  As  a  result 
there  are  a  multitude  of  distracting  details  in  range  of  our  eyes 
at  one  time  and  we  regard  a  moving  picture  with  an  unwinking 
intentness  which  is  very  rarely  called  for  in  ordinary  life. 

Neither  do  the  extrinsic  muscles  of  the  eyes  escape  strain. 
Even  if  the  picture  were  an  unchanging  one  the  mere  matter  of 
watching  it  implies  the  co-ordination  of  the  eyes  which  depends  on 
the  extrinsic  muscles.  But  in  the  moving  picture  motions  which 
are  intended  to  appear  to  us  smooth  and  continuous  as  they  were 
in  life  really  proceed  by  a  very  rapid  succession  of  stops  and 
starts  and  the  actions  are  usually  completed  in  a  very  much  shorter 
time  than  they  took  in  real  life.  Possibly  by  being  less  economical 
in  the  matter  of  film  the  jerky  motion  of  the  figures  in  the  picture 
could  be  largely  done  away  with  but  in  practice  we  are  always 
conscious  of  and  instinctively  try  to  follow  these  movements.  As 
a  result  our  extrinsic  eye  muscles  are  in  a  state  of  steady  nervous 
and  muscular  tension  which  is  both  abnormal  and  fatiguing. 

But  the  most  obvious  cause  of  eye-strain  in  watching  moving 
pictures  are  the  rapid  jerky  motions  of  the  whole  picture,  the 
occasional  showers  of  light  flashes,  and  the  variations  in  light 
intensity  which  give  us  a  sensation  of  flickering  light.  The  first 
two  are  said  to  be  due  to  imperfections  in  the  film  and  can  be 
done  away  with  by  more  careful  construction,  but  the  flicker 
seems  to  be  inherent  in  the  business.  I  should  say  that  the  dis- 
agreeable effect  of  flicker  was  only  partly  due  to  its  effect  on  the 
iris.  Light  falling  on  the  retina  causes  a  stimulation  of  a  centre 
in  the  brain  which  in  turn  causes  a  contraction  of  the  pupil.  If 
the  light  remains  steady  the  retina  becomes  adapted  to  it  and  the 
pupil  dilates  somewhat.  When  the  light  stimulus  is  removed  the 
pupil  dilates  pretty  widely.  This  whole  process  requires  an 
appreciable  time  and  the  degree  of  contraction  depends  not  only 
on  the  brightness  of  the  light  to  which  the  eye  is  exposed  but  also 
on  the  degree  of  illumination  to  which  it  had  previously  become 
adapted.  So  that  a  light  of  rather  low  intensity,  provided  the 
4 


192  TRANSACTIONS  I.  E.  S. — PART  II 

room  is  comparatively  dark,  may  cause  marked  contraction.  But 
the  flicker  of  the  moving  picture  applies  and  removes  the  stimulus 
of  bright  light  much  more  rapidly  than  the  muscle  can  respond, 
and  as  a  result  the  pupil  remains  in  a  state  of  rapid  oscillation 
which  is  not  synchronous  with  the  actual  variations  in  light  inten- 
sity. The  outcome  is  not  only  fatigue  of  the  sphincter  pupillae 
but  a  state  of  nerve  exhaustion  from  the  too  rapid  discharge  of 
nerve  impulses. 

Another  difficulty  which  has  to  do  with  the  adaption  of  the  eye 
to  light  is  this.  All  observers  are  agreed  on  the  retinal  fatigue 
that  comes  from  having  one  portion  of  the  retina  exposed  to  light 
while  the  rest  of  it  is  in  comparative  darkness.  A  strong  in- 
candescent light  in  a  dark  room  is  too  bright  to  look  directly  at 
while  if  the  room  be  filled  with  daylight  the  bulb  seems  hardly 
visible.  In  the  moving  picture  one  is  looking  intently  at  a  screen 
on  which  the  light  may  not  be  intrinsically  very  intense  but,  by 
comparison  with  the  surroundings,  seems  very  bright  indeed.  It 
would  probably  be  very  much  easier  for  the  eyes  of  the  observers 
if  the  theatres  were  lighted  in  such  a  way  that  while  no  direct  light 
fell  on  the  eyes  the  general  illumination  was  as  good  as  was 
compatible  with  visibility  of  the  screen. 

Mr.  F.  A.  Vaughn  (communicated)  :  The  author's  discussion 
of  eye-strain,  discomfort  and  permanent  injury  caused  by  the 
motion  pictures  would,  it  is  thought,  lead  one  to  suppose  that  it 
was  a  matter  which  could  be  dismissed  without  very  great  con- 
cern, as  the  troubles,  when  apparent  at  all,  are  caused,  according 
to  the  author,  by  undue  "unsteadiness  of  the  picture  and 
blemishes  of  the  film  used".  While  the  discussion  may  be  accurate 
as  far  as  it  covers  the  case,  it  is  believed  that  it  should  be  pointed 
out  that  there  have  been  severe  cases  of  ocular  disturbances  cited 
by  opthalmologists  and  the  subject  is  of  enough  concern  to  have 
received  attention  in  the  Calif ornia  State  Journal  of  Medicine  in 
August,  1912,  under  the  subject  of  "Ocular  Disturbances  caused 
by  the  Cinematograph"  through  the  authorship  of  Morton  E. 
Hart,  M.  D.,  San  Francisco.  Dr.  Hart  discusses  it  in  part  as 
follows : 

Ocular  disturbances  due  to  the  cinematograph  have,  up  to  the  present 
time,  received  practically  no  mention  in  medical  literature.  It  seems 
strange  that  this  should  be  the  case,  for  no  doubt  it  has  fallen  to  the 


ILLUMINATION   OF   MOTION    PICTURE   PROJECTORS  I93 

lot  of  almost  every  oculist,  particularly  in  the  large  cities  to  have  seen 
and  treated  many  patients  suffering  from  this  new  disease.  And  there 
are  very  good  reasons  that  there  should  be  ocular  disturbances  from  this 
new  plaything  of  the  peopie.     *     *     * 

The  ocular  disturbances,  classified  under  the  generic  term  of  "cine- 
matopthalmia,"  are  really  disturbances  of  vision  due  to  traumatism,  and 
are  matters  of  degree.  The  process  is  the  same  in  all  of  the  conditions. 
There  are  those  cases  which  are  merely  transient  in  their  disturbance. 
When  the  picture  is  first  thrown  on  the  screen,  the  individual  is  incon- 
venienced by  photophobia  and  a  few  tears.  He  closes  his  eyes  and  these 
symptoms  soon  pass  away  after  a  few  seconds  of  repose,  and  the  retina 
accustoms  itself  to  the  new  condition  of  affairs.  A  further  degree  is  of 
longer  duration;  the  retina  cannot  adopt  itself  to  the  fatigue  imposed  on 
it  and  each  time  the  individual  opens  his  eyes,  the  symptoms  reappear. 
It  is  impossible  to  continue  the  spectacle.  After  leaving  the  theatre,  the 
disturbance  still  persists  and  in  addition  to  the  mild  photophobia  and 
lacrymation  there  ensues  a  slight  reddening  of  the  conjunctiva.  A  few 
hours,  or  at  least  a  night's  rest,  will  return  the  eyes  to  their  normal  tone. 

In  the  third  degree  of  disturbance,  the  symptoms  are  more  severe 
and  the  return  to  the  normal  somewhat  prolonged.  Here  the  phtophobia, 
lacrymation  and  conjunctivitis  persist  for  several  days  and  in  addition, 
we  have  a  smarting  and  itching  of  the  eyes. 

In  the  very  severe  cases,  besides  the  inflammation  of  the  conjunctiva 
with  its  attendant  symptoms  of  lacrymation  and  phtophobia,  we  have 
very  definite  asthenopic  symptoms,  both  accommodative  and  retinal — the 
former  due  to  the  ciliary  strain  and  the  latter  due  to  a  hyperesthesia  of 
the  retina.  The  distant  vision  remains  normal.  Under  examination  these 
patients  are  found  to  have  no  error  of  refraction  or  lesion  of  the  fundus. 
A  case  in  question  may  here  be  cited : 

E.  R.,  female,  age  16,  was  brought  to  me  with  the  following  complaint  :  Eyes  burned 
and  itched  and  the  lids  were  red,  particularly  at  night.  Reading  was  impossible  on  ac- 
count of  blurring  of  the  page.  No  headaches.  This  condition  would  clear  up  after  a 
night's  rest,  to  reappear  again  at  frequent  intervals. 

On  examination  a  slight  reddening  of  the  conjunctiva  was  found  and  under  a  myd- 
riatic an  error  of  one  degree  of  hyperopia,  which  was  corrected.  The  near  point  was  nor- 
mal, showing  no  error  of  accommodation.  Of  course  this  was  tested  before  using  the 
mydriatic.  No  lesion  of  the  fundus  was  found.  Unfortunately  the  patient  could  not  be 
seen  during  an  attack. 

After  wearing  the  glasses  for  several  weeks,  the  patient  reported,  stating  that  the  con- 
dition had  not  improved.  She  was  then  closely  questioned  and  it  was  found  that  it  was 
her  habit  to  attend  a  moving  picture  show  at  least  four  times  a  week  after  school  and 
unbeknown  to  her  mother.  She  was  forbidden  this  amusement  and  the  condition  entire- 
ly cleared  up. 

Fortunately  these  ocular  disturbances  are  not  serious  and  will  clear 
up  under  simple  collyria  and  rest. 

The  question  will  naturally  arise,  how  can  we  do  away  with  the  cause 
of  the  trouble? 

First:     The  films  must  be  perfect  and  free  from  all   imperfections. 


194  TRANSACTIONS  I.  Z.  S. — PART  II 

We  have  all  noticed  the  scratches  on  the  pictures,  particularly  at  the  end 
of  the  reels,  due  to  careless  handling.  When  we  realize  that  the  average 
picture  thrown  on  the  screen  is  about  97,000  times  larger  than  the  original 
size  of  the  individual  film,  we  can  appreciate  that  even  the  smallest 
blemish  on  the  films  will  be  tremendously  magnified  on  the  curtain  and 
will  have  a  correspondingly  bad  effect  on  the  eyes. 

Second :  The  illumination  must  be  steady,  must  not  vary  and  must 
neither  be  too  bright  nor  too  dim,  for  this  causes  fatigue. 

Third :  The  speed  with  which  the  films  are  turned  must  be  regular. 
Any  irregularity  will  have  a  tendency  to  cause  ocular  fatigue. 

Fourth :  The  position  of  the  spectator  is  very  important  and  should 
receive  proper  regulation  at  the  hands  of  the  authorities.  First  of  all, 
there  should  be  no  seats  placed  at  the  sides  of  the  auditorium.  Every 
seat  should  be  in  direct  line  with  the  curtain.  This  will  do  away  with 
the  distortion  of  the  picture.  Anyone  who  has  had  the  experience  of 
sitting  on  the  side,  can  appreciate  the  intense  strain  and  fatigue  placed 
on  the  eyes. 

No  seat  should  be  placed  nearer  than  twenty  feet  from  the  screen 
and  further  if  practicable,  depending  upon  the  size  of  the  picture  on  the 
curtain.  This  will  do  away  with  any  accommodative  effort  on  the  part  of 
the  spectator  and  thus  will  reduce  the  fatigue  to  a  minimum.  The  nearer 
the  screen  the  greater  the  fatigue,  so  the  seats  at  the  rear  of  the  audi- 
torium are  the  best. 

Some  people  are  very  susceptible  to  the  influence  of  motion 
pictures,  not  being  able  to  sit  through  one  performance  without 
headache  or  severe  eye-strain.  It  will  thus  be  seen  that  quite 
painful  and  somewhat  serious  cases  of  ocular  disturbances  have 
been  caused  and  probably  will  be  caused  in  the  future  by  motion 
pictures  and  that,  while  this  may  be  due  to  imperfections  in  their 
operation  or  presentation,  these  imperfections  will  probably  obtain 
in  the  future  as  well,  and  this  subject  should  therefore  receive 
serious  consideration  from  ophthalmologists,  as  well  as  illuminat- 
ing engineers,  at  least  until  the  imperfections  are  eliminated. 
After  that  it  may  be  determined  whether  there  is  still  remaining 
serious  objection  to  them  from  the  standpoint  of  ocular  comfort. 
Dr.  P.  W.  Cobb  (communicated)  :  The  motion  picture  man 
has  a  set  of  problems  pretty  much  his  own.  Mr.  Martin's  paper 
touches  some  of  them  in  a  way  which  interests  me.  I  am  sur- 
prised at  his  statement  that  the  illumination  on  the  screen  is  only 
0.4  foot-candle  but  such  surprises  are  many  for  one  who  attempts 
to  estimate  illumination  from  visual  impression.  The  motion 
picture  auditorium  has  to  have  a  low  general  illumination,  just 


ILLUMINATION   OE   MOTION   PICTURE   PROJECTORS  I95 

enough  for  a  person  whose  eyes  are  light-adapted  on  his  entrance 
to  the  theatre  to  see  to  take  a  place.  Mr.  Martin  places  this  at  an 
average  of  0.2  foot-candles.  This  sets  the  pace,  as  it  were,  for 
the  eye  and  determines  the  illumination  necessary  upon  the  screen 
in  order  to  make  the  pictures  stand  out  brilliantly.  The  question 
of  contrast  enters  here  as  an  important  factor;  the  contrast  be- 
tween the  picture  on  the  screen  on  the  one  hand,  and  on  the  other 
its  surroundings,  the  seats,  walls  and  ceiling  of  the  theatre.  The 
surface  brightness  of  these  latter  is  a  factor  in  determining  the 
visual  brightness  of  the  screen,  equal  in  importance  to  the  lumens 
per  unit  area  of  the  screen  itself. 

Speaking  of  contrast,  I  should  like  to  ask  if  tinting  the  general 
illumination  of  the  room  has  ever  been  used  to  modify  the  appar- 
ent color  of  the  screen.  For  instance,  I  should  expect  that  by 
tinting  the  lamps  in  the  auditorium  a  bluish  color,  the  pictures 
could  be  made  to  appear  yellowish,  or  by  making  the  lamps 
greenish  the  screen  would  show  a  tendency  to  a  rosy  pink.  A 
small  amount  of  tint  in  the  general  illumination  would  be  enough, 
I  should  think,  to  get  a  pronounced  contrast  effect  on  the  screen. 

It  is  gratifying  to  note  that  the  question  of  eye-strain  has  been 
considered.  At  0.4  foot-candles  the  question  of  too  intense  light 
cannot  merit  discussion  and  glare  can  be  ruled  out.  What  should 
be,  and  is,  considered  is  flicker  and  unsteadiness  of  the  image. 

I  cannot  agree  with  what  Mr.  Martin  says  as  to  the  effect  of 
flicker  being  confined  to  the  muscle  of  the  iris  which  controls  the 
pupil. 

It  has  been  shown  that  when  the  eye  is  kept  in  the  dark  and 
illuminated  by  a  momentary  flash  of  light  the  reaction  of  the 
pupil  is  delayed  for  something  like  half  a  second.  The  con- 
traction follows  and  the  pupil  does  not  begin  to  dilate  until  a 
lapse  of  about  10  seconds.  Such  movements  are  altogether  too 
slow  to  follow  the  flicker  of  a  motion  picture  which  has  a  rate 
say  of  10  per  second  or  thereabouts.  As  a  matter  of  fact  the 
pupil  is  always  undergoing  slight  fluctuations  in  size  in  the 
absence  of  all  changes  in  light,  a  fact  which  can  be  verified  by 
anyone  who  cares  to  examine  his  own  pupil  by  means  of  a  mirror 
and  a  lens  of  low  power.  It  would  appear  to  me  that  the  examina- 
tion of  the  pupil  under  strongly  flickering  light  would  be  attended 


I96  TRANSACTIONS  I.  E.  S. — PART  II 

with  great  uncertainties  and  make  it  difficult  to  estimate  the 
amount  of  fluctuation  and  compare  it  with  the  normal. 

Nevertheless,  flicker  is  undoubtedly  disturbing  to  the  eyes,  a 
fact  which  was  much  made  use  of  in  getting  electric  incandescent 
lights  into  use  when  they  were  first  introduced  to  the  public.  The 
exact  way  in  which  flicker  embarrasses  the  eyes  is,  I  believe,  still 
not  known. 

By  far  the  most  serious  evil  of  motion  pictures  seems  to  me  to 
be  unsteadiness  of  the  image,  which  I  am  glad  to  know  is  an  un- 
necessary one.  A  person  leaving  a  motion  picture  theatre  to  go 
back  to  an  occupation  not  especially  involving  the  eyes  may,  as 
Mr.  Martin  says,  find  the  difficulty  which  he  experienced  while 
viewing  the  pictures  to  disappear  rapidly,  if  he  continues  to 
notice  it  at  all.  But  it  must  be  far  different  with  one  whose  call- 
ing depends  on  close  application  of  the  eyes.  Here  the  muscles 
of  the  eye,  which  have  been  subject  to  all  sorts  of  surprises  in 
trying  to  fix  upon  a  wavering  image  are  taken  away  from  this 
only  to  do  more  work  of  a  similar  kind,  reading,  sewing  or  what 
not,  which  calls  for  the  finest  sort  of  co-ordination  of  the  twelve 
extremal  muscles  of  the  eyes.  When  one  reflects  upon  the 
delicate  balance  of  these  muscles,  and  the  incessant  and  rapid 
movements  which  they  must  accurately  carry  out  in  fine  eye-work 
it  seems  impossible  that  such  a  shaking  up  as  they  get  in  attempt- 
ing to  fix  an  image  which  jerks  about  irregularly  upon  the  screen 
should  not  unfit  the  eyes  to  a  greater  or  less  extent  for  subsequent 
work.  It  is  therefore  good  to  hear  that  such  pictures  need  not 
be,  and  to  feel  that  they  will  soon  be  eliminated  from  the  motion 
picture  world. 

As  to  the  suggestion  that  observations  be  made  upon  motion 
picture  operators  to  detect  changes  in  the  eyes  due  to  the  pictures 
I  wish  to  ask  does  the  operator  himself  watch  the  pictures  at  all 
attentively?  My  feeling  is  that  he  does  not,  or  at  most,  watches 
them  "out  of  the  corner  of  his  eye"  only  to  detect  gross  defects 
in  the  film  or  in  the  action  of  the  machine.  It  is  an  open  question 
in  my  mind  whether  he  would  suffer  as  much  in  a  week  as  the 
casual  visitor  to  his  theatre  does  in  an  hour. 

The  visitor  is  eager  and  intensely  interested  in  the  story  of  the 
picture  (which  is  a  very  old  story  to  the  operator)  and  applies  his 


ILLUMINATION  OF   MOTION    PICTURE;   PROJECTORS  I97 

attention  and  his  eyes  closely  to  the  image.  It  is  just  this  close 
attention  which  whips  up  the  eye-muscles  to  the  task  of  observing 
the  image  and  spurs  them  on  to  their  own  confusion  if  the  latter 
be  unsteady. 

Mr.  Edward  L.  Simon:  (communicated)  :  This  paper  is  in- 
deed food  for  extended  investigation  and  the  time  is  now  ripe 
for  a  uniform  and  scientifically  designed  and  arranged  source  of 
light  for  projection. 

The  large  film  manufacturers  have  and  are  expending 
hundreds  of  thousands  of  dollars  for  the  improvement  of  the  films 
and  these  films  when  delivered  to  the  exhibitors  are  mechanically 
perfect.  This  does  not  apply  to  the  small  manufacturers  and  the 
audience  is  not  able  to  judge  between  a  good  and  bad  film  owing 
to  a  defective  projecting  machine. 

The  projecting  machine  of  the  future  must  be  a  presicion 
machine  mounted  on  a  concrete  or  some  other  rigid  foundation. 
The  light  should  be  a  cold  light  and  to  get  rid  of  the  flicker  the 
screen  should  never  be  dark,  which  is  now  the  case  sixteen  times 
every  second.  It  is  not  right  that  a  beautiful  picture  should  be 
spoiled  in  its  projection  by  being  run  through  a  loose  and  untrue 
projecting  machine  with  a  projecting  light  that  is  operated 
according  to  the  individual  operators  ideas.  It  will  not  be  long 
before  the  manufacturers  will  be  compelled  to  insist  upon  rigid 
enforcements  in  the  way  of  perfect  projecting  machines  and 
proper  light  source. 

The  film  as  used  to-day  is  enlarged  one  hundred  and  fifty 
diameters.  The  large  productions  in  the  future  will  not  be  en- 
larged more  than  fifty  diameters  and  the  audience  will  have  the 
pleasure  of  looking  at  pictures  as  steady  as  stereoptican  views. 

Mr.  Louis  C.  Smith  (communicated)  :  In  my  experience  in 
projecting  pictures,  I  have  found  that  considerable  attention 
should  be  paid  to  the  regulating  of  the  quantity  of  light  for 
different  films,  slides,  etc.  I  do  not  apply'  this  to  dense  films 
entirely,  but  to  various  films  of  varying  subjects.  Recently,  I 
projected  some  cloud  scenes,  the  effect  of  which  when  lighted 
with  all  the  available  energy  was  so  contrasty  that  it  was  un- 
natural. In  general,  though,  I  have  found  that  the  general  public 
like  the  bright  contrasty  photographs  and  motion  pictures.     More 


I98  TRANSACTIONS  I.  K.  S. — PART  II 

attention  seems  to  be  paid  to  the  theme  of  the  picture  than  the 
details  of  it.  This  is  evidently  so  partly  because  defects  in  the 
pictures  are  so  usual. 

Although  large  amounts  of  money  are  spent  in  the  moving 
picture  field,  the  theatres  seldom  get  good  complete  projection 
outfits ;  then  again  the  operator  usually  does  little  more  than  turn 
the  crank  mechanically;  he  pays  little  attention,  if  any,  to  getting 
the  best  results  out  of  his  apparatus. 

There  are  subjects  which  require  great  speed  and  others  that 
should  not  be  projected  so  rapidly;  usually  all  are  projected 
alike.  Many  things  depend  directly  upon  the  taste  and 
experience  of  the  operator. 

Some  of  the  most  pleasing  slides  have  been  projected  by  a 
calcium  light.  This  seems  to  be  due  to  the  yellowness  that  the 
author  speaks  of.  No  doubt  a  yellowish  bright  light  would  be 
suitable  for  many  films. 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 

Published  monthly,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIII 


MAY.  1913 


No.  5 


Council  Notes. 

A  meeting  of  the  Council  was  held  in 
the  general  offices  of  the  society,  29 
West  39th  Street,  New  York,  May  9, 
101 3.  Those  in  attendance  were:  Pres- 
ton S.  Millar,  president ;  George  S. 
Barrows,  C.  O.  Bond,  J.  W.  Cowles, 
Joseph  D.  Israel,  general  secretary;  A. 
E.  Kennelly,  Xorman  Macbeth,  L.  B. 
Marks,  treasurer;  W.  Cullen  Morris, 
W.  J.  Serrill,  R.  C.  Ware  and  Arthur 
Williams. 

The  Executive  Committee  reported 
that  it  had  held  a  meeting  April  25. 
The  business  transacted  by  the  com- 
mittee is  given  in  the  following  minutes 
of  the  meeting: 

It  was  decided  to  accept  the  invitation  of  the 
Pittsburgh  section  to  hold  the  1913  convention 
of  the  society  at  Pittsburgh  in  the  fall. 

After  a  general  discussion  of  convention 
plans  President  Millar  was  directed  to  appoint 
a  general  convention  committee.  It  was  under- 
stood that  the  general  committee  might  appoint 
such  sub-committees  as  may  be  required  to  care 
for  the  various  details  and  arrangements  of  the 
convention. 

A  draft  of  the  society's  conspectus,  submitted 
by  Mr.  Macbeth,  was  discussed  informally. 

The  above  report  was  received  and 
concurred  in  by  the  Council. 

Vouchers  1273  to  1314  inclusive, 
covering  May  bills  aggregating  $1,086.70, 
which  had  been  approved  by  the  Finance 
Committee  were  authorized  paid.  A 
monthly  report  on  the  society's  finances 
and  membership  was  received  from  the 


general  secretary.  The  expenses  of  the 
first  four  months  of  1913  were  said  to 
have  totaled  $2,823.29.  The  membership 
including  the  additions  and  defections 
presented  at  the  meeting  was  said  to 
have  totaled  1,363  members.  The  mem- 
bership at  the  beginning  of  the  year 
was   1,335. 

Reports  on  section  activities  were 
received  from  the  following  vice-presi- 
dents:  W.  J.  Serrill,  Philadelphia;  J. 
\V.  Cowles,  New  England ;  Norman 
Macbeth,  New  York;  Howard  S.  Evans, 
Pittsburgh,  and  James  Cravath,  Chicago. 

The  Section  Development  Committee 
reported  that  it  had  completed  a  guide 
on  section  management  which  would 
be  issued  shortly. 

The  chairman  of  the  Committee  on 
Papers  reported  tentatively  on  the  pro- 
gram of  papers  for  the  1913  convention. 

A  brief  and  informal  report  on  the 
progress  of  plans  for  the  1913  Conven- 
tion was  received  from  President  Millar. 
The  following  appointments  of  the 
president  to  the  Convention  Executive 
Committee  were  approved :  C.  A.  Little- 
field,  chairman  ;  M.  C.  Rypinski,  D.  Mc- 
Farlan  Mo'ore,  O.  H.  Fogg  and  Thomas 
S.  Henderson. 

Additional  appointments  to  committees 
made  by  the  president  since  the  last 
Council  meeting  were  approved. 

The  following  committee  of  five 
tellers     was     appointed     to     count    the 


TRANSACTIONS    I.    K.    S. — PART    I 


ballots  of  the  annual  election:  G.  B. 
Nichols,  chairman ;  L.  J.  Lewinson,  A. 
L.  Powell,  Raymond  W.  Stafford  and 
Thomas    Scofield. 

Mr.  L.  B.  Marks  presented  an  oral 
and  informal  report  of  the  progress  of 
the  plans  of  the  Gas  Congress  which  is 
to  be  held  in  San  Francisco  in  1915.  Mr. 
Marks  is  the  representative  of  the 
society  on  the  committee  which  is  to 
plan  the  Congress. 

Mr.  L.  B.  Marks  chairman  of  the 
Committee  on  Factory  Lighting  Legisla- 
tion reported  that  Governor  Sulzer  of 
the  State  of  New  York  had  signed  on 
April  17,  Bill  No.  26  which  is  entitled 
"An  act  to  amend  the  labor  law  in  rela- 
tion to  protection  of  employees  operat- 
ing machinery,  thus  creating  machinery 
and  the  lighting  of  factories  and  work 
rooms."  A  section  of  this  bill  relating 
particularly  to  the  lighting  of  factories, 
passage-ways  and  work-rooms  was 
drafted  in  accordance  with  recommenda- 
tions made  by  the  committee.  This  sec- 
tion of  the  bill  was  printed  in  the  April 
issue  of  the  Transactions. 

Dr.  A.  E.  Kennelly  reported  orally 
for  the  Committee  on  Nomenclature  and 
Standards.  He  said  that  a  written  re- 
port would  be  forthcoming  from  the 
committee   shortly. 


Section  Notes. 

CHICAGO  SECTION. 

At  a  meeting  of  the  Chicago  Section 
in  the  auditorium  of  the  Western  So- 
ciety of  Engineers,  May  14,  Mr.  M. 
Luckiesh  of  the  National  Electric  Lamp 
Association,  Cleveland,  gave  a  lecture 
on  "Light  and  Art."  The  lecture,  which 
was  accompanied  by  a  series  of  well  plan- 
ned demonstrations  showing  the  effect 
of  direction,  and  quantity  and  quality  of 


light  on  objects  of  art  and  design, 
proved  to  be  intensely  interesting  to  the 
forty  members  and  guests  in  attendance. 
The  lecture  had  been  given  before  meet- 
ings of  several  other  sections  of  the 
society. 

At  the  June  meeting,  which  will  prob- 
ably be  held  on  the  18th,  Mr.  Arthur 
J.  Sweet  will  present  a  paper  on  car 
lighting. 


NEW    ENGLAND  SECTION. 

The  May  meeting  of  the  New  Eng- 
land Section  was  postponed.  No  date 
or  program  has  been  scheduled  for  a 
June  meeting. 

NEW  YORK  SECTION. 

Before  a  largely  attended  meeting  of 
the  New  York  Section,  May  8,  Mr. 
Bassett  Jones,  Jr.,  gave  a  very  interest- 
ing lecture  and  series  of  demonstrations 
on  theater  lighting.  The  meeting  was 
held  in  the  Clymer  Street  Theater, 
Brooklyn,  through  the  courtesy  of  the 
proprietors.  The  uusal  monthly  dinner 
at  Keene's  Chop  House,  preceded  the 
meeting. 

PHILADELPHIA   SECTION. 

The  May  meeting  of  the  Philadelphia 
Section,  in  the  Franklin  Institute,  May 
16,  was  addressed  by  Dr.  Herbert  E. 
Ives  on  "Some  Home  Experiments  in 
Illumination  from  Large  Light  Sources," 
and  Mr.  C.  A.  Peterson  on  "The  Design 
of  Combination  Fixtures."  There  was 
also  an  exhibition  of  the  latest  types  of 
residential  glassware. 

The  Section  will  have  a  June  outing 
which  will  be  a  joint  outing  with  the 
Philadelphia  Section  of  the  American 
Institute  of  Electrical  Engineers  on  the 
grounds  of  the  Philadelphia  Electric 
Company      Athletic      Association,      on 


TRANSACTIONS    I.    E.    S. — PART    I 


Saturday  afternoon,  June  "th.  There 
will  be  a  baseball  game  between  mem- 
bers of  the  Illuminating  Engineering 
Society  and  the  American  Institute  of 
Electrical  Engineers.  Prof.  Franklin  of 
Lehigh  University  will  give  a  lecture 
on  "Baseball  Curves."  Supper  will  be 
served  on  the  grounds 

On  June  20  the  section  will  hold  a 
short  business  meeting  and  dinner  at 
the  Engineers'  Club  instead  of  a  regular 
meeting.  There  will  be  some  short 
speeches  but  no  paper. 

PITTSBURGH   SECTION. 

A  meeting  of  the  Pittsburgh  Section 
was  held  May  16.  Mr.  J.  L.  Minick  of 
the  Pennsylvania  Railroad  Company  read 
a  paper  on  "Passenger  Car  Lighting." 
The  paper  appears  in  this  issue  of  the 
Transactions. 

For  the  June  meeting  Messrs.  Ward 
Harrison  and  E.  J.  Edwards  of  the 
National  Electric  Lamp  Association  are 
preparing  a  paper  on  the  lighting  of  the 
new  office  buildings  of  the  National 
Electric  Lamp  Division  of  the  General 
Electric  Company,  at  Nela  Park,  Cleve- 
land. 


New  Members. 

The  following  applicants  were  elected 
members  of  the  society  at  a  meeting  of 
the  Council,  May  9,  1913 : 

Amkine,  T.  H. 

General    Electric    Company,    Harri- 
son.  N.  J. 

Bergman,  Rudolph  L. 

Salesman,  Benjamin   Moore  &  Co., 

231  Front  Street,  Brooklyn,  N.  Y. 
Bern  hard.,  Albert  H. 

1070     Bedford     Avenue,     Brooklyn, 

N.  Y. 


Bryant,    John    Myron 

Asst.  Professor  of  Electrical  En- 
gineering, University  of  Illinois, 
Urbana,  111. 

Davidson,  John  M. 

Civil  and  Sanitary  Engineer,  Ameri- 
can Sheet  and  Tin  Plate  Company, 
1224  Frick  Building,  Pittsburgh,  Pa. 

Ehrlich,    Howard 

Associate  Editor,  Electrical  Re- 
view and  Western  Electrician,  608 
S.  Dearborn  Street,  Chicago,  111. 

Ellis,  Edgar  J. 

President  United  Electric  Const. 
Company,  1727  Sansom  Street, 
Philadelphia,  Pa. 

Frith,  Andrew  M. 

Burnham-Frith  Electric  Company, 
MacDougall  Avenue,  Edmonton, 
Alta.,  Can. 

Froelich,  J.  M. 

Arc  Lighting  Engineer,  Duquesne 
Light  Company,  435  Sixth  Avenue, 
Pittsburgh,   Pa. 

Gray,  A.  A. 

Managing  Editor,  Electrical  Re- 
z'iezu  and  Western  Electrician,  608 
S.  Dearborn  Street,  Chicago.  111. 

Grondahl,  L.  O. 

Instructor,  Carnegie  Institute  of 
Technology,  Pittsburgh,  Pa. 

Henderson,  R.  G. 

District  Manager,  General  Electric 
Company,  30  Church  Street,  New 
York,  N.  Y. 

Ives,   Arthur    S. 

Ives  &  Davidson,  84  William  Street, 
New  York,  N.  Y. 

Kaulke,  Johannes 

Electrical  Engineer,  General  Elec- 
tric Company,  Sussex  and  Fourth 
Streets,  Harrison,  N.  J. 

Kelley,  J.  B. 

Salesman,  Frank  H.  Stewart  Elec- 
tric Company,  35  N.  7th  Street, 
Philadelphia,    Pa. 


TRANSACTIONS    I.    E.    S. — PART    I 


Kelly,  Clarence  B. 

Chief  Estimator;,  United  Electric 
Const.  Co.,  1708  Sansom  Street, 
Philadelphia,  Pa. 

Latta,  J.  E. 

Associate  Editor,  Electrical  Re- 
view, 608  S.  Dearborn  Street, 
Chicago,    111. 

Landerdale,  Jesse  E. 

Sales  Engineer,  National  X-Ray 
Reflector  Company,  235  West 
Jackson  Boulevard,  Chicago,  111. 

Mason,  Frank  L. 

Instructor,  Dept.  Electrical  En- 
gineering, Columbia  University, 
New  York,  N.  Y. 

Mauser,  R.  H. 

Engineer's  Assistant,  Consolidated 
Gas  Company,  124  East  15th  Street, 
New  York,  N.  Y. 

McNeil,  R.  S. 

General  Electric  Company,  Harri- 
son, N.  J. 

Otto,  William  G. 

General  Sales  Manager,  Walker 
Electric  Company,  2338  Nobb  Street, 
Philadelphia,  Pa. 

Rowland,  Ernest  W. 

Chief  Bond  Inspector,  Public  Ser- 
vice Railway,  Newark,  N.  J. 

Rypinski,  M.  C. 

Manager  Detail  and  Supply  Dept, 
Westinghouse  Electric  and  Mfg. 
Company,  163  Broadway,  New 
York,  N.  Y. 

Truitt,  Thomas  Gibe 

Imperial  Electric  Company,  1022 
Arch  Street,  Philadelphia.  Pa. 

Watkins,  Howard  E. 

Vice-President  and  Designer,  The 
Enos  &  Watkins  Co.,  36  West  37th 
Street,  New  York,  N.  Y. 


Sustaining   Member. 

The  United  Electric  Light  and  Power 
Company,  New  York,  was  elected  a  sus- 
taining member  of  the  society  at  a 
meeting  of  the  Council,  Ma}'  9,  1913. 


Joint  Meeting  I.  E.  S.  and  A.  G.  I. 

The  following  announcement  has 
been  issued  by  a  special  committee  of 
the  society,  which  is  arranging  for  a 
joint  meeting  with  the  American  Gas 
Institute : 

"A  joint  session  of  the  American  Gas 
Institute  and  the  Illuminating  Engineer- 
ing Society,  will  be  held  during  the  third 
week  in  October  at  the  annual  meeting 
of  the  Institute  in  Richmond,  Ya. 
While  it  is  true  that  a  number  of  gas 
men  are  active  and  influential  in  the 
work  of  our  society,  still  a  regret  is  fre- 
quently expressed  that  gas  men  and  gas 
companies  generally,  do  not  take  a 
more  active  interest.  The  gas  business 
is  so  intimately  bound  up  with  our  pur- 
poses and  objects  that  it  seems  un- 
necessary to  dwell  on  the  value  of  our 
society  to  that  great  industry. 

Your  committee  believes  that  this 
joint  meeting  will  furnish  an  excellent 
opportunity  for  the  arousing  and 
stimulation  of  such  an  interest. *  It 
urges  on  the  membership  of  our  society 
the  importance  of  a  large  attendance  at 
this  joint  meeting,  and  of  an  abundant 
discussion  of  the  papers  presented  to  it. 

We  are  notified  by  the  Committee  on 
Papers  that  the  following  papers  have 
been  secured : 

"Some  phases  of  the  Illumination  of 
Interiors,"  by  Preston  S.  Millar,  Secre- 
tary, Electrical  Testing  Laboratories, 
New  York,  N.  Y.  A  lecture-demonstra- 
tion employing  minature  rooms  to 
illustrate    several   well  known   types  of 


TRANSACTIONS    I.    E.    S.—  PART    I 


lighting  installations,  indicating  their 
peculiarities  and  good  and  bad  features. 

"The  Importance  of  Direction,  Qual- 
ity and  Quantitative  Distribution  of 
Light  in  Illumination,"  by  M.  Luckiesh, 
assistant  physicist,  National  Electric 
Lamp  Association,  Cleveland,  Ohio,  a 
lecture-demonstration  chiefly  by  sub- 
jects taken  from  the  fine  arts. 

"Gas  Lighting  of  Interiors,"  by  C.  A. 
Luther,  Illuminating  Engineer,  Peoples 
Gas  Light  and  Coke  Company,  Chicago. 
111.,  a  paper  dealing  with  the  lighting  of 
interiors  by  gas,  and  illustrating  the 
manner  in  which  gas  is  used  to  obtain 
the  results  explained  in  Mr.  Millar's 
demonstration. 

"Street  Lighting  by  Gas,''  by  F. 
V.  Westermaier,  Engineer,  Welsbach 
Street    Lighting   Company  of   America, 


Philadelphia,  Pa.,  a  paper  dealing  with 
street  lighting  from  the  standpoint  of 
the  most  modern  methods. 

This  will  evidently  be  an  entertaining 
and  instructive  session.  The  date  of 
the  joint  meeting  will  be  announced 
later. 


1913  I.  E.  S.  Convention. 

The  seventh  annual  convention  of 
the  Illuminating  Engineering  Society 
will  be  held  in  Pittsburgh,  September 
22  to  26  inclusive.  A  committee  of 
which  Mr.  C.  A.  Littlefield,  55  Duane 
Street,  Xew  York,  is  chairman,  is  en- 
gaged in  formulating  plans  for  the 
biggest  and  most  successful  convention 
yet  held  by  the  society. 


TRANSACTIONS 


OF  THE 


Illuminating 
Engineering  Society 


MAY,  1913 


PART  II 


ck 


<K 


Papers,  Discussions  and  Reports 


[  MAY,  1913  ] 

CONTENTS  -  PART  II 


Street  Lighting  of  Greater  New  York.     By  C.  F.  La- 
combe  199 

Illumination  of  Passenger  Cars.     By  J.  L.  Minick 214 


H 


STREET  LIGHTING  OF  GREATER  NEW  YORK.* 


BY  C.  F.  LACOMBE. 


Synopsis:  Included  in  the  following  paper  is  a  description  of  the  plan 
or  system  of  street  lighting  of  the  City  of  New  York.  With  its  accom- 
panying illustrations  the  paper  presents  a  comprehensive  outline  of  the 
progress  which  has  been  made,  particularly  within  the  past  decade. 

In  appearing  before  you  this  evening  in  an  endeavor  to  de- 
scribe the  street  lighting  of  the  City  of  New  York  and  the  efforts 
made  to  improve  it,  and  to  make  it  attractive  and  artistic,  my 
attitude  is  not  that  of  a  Philistine,  satisfied  fully  with  the  engi- 
neering and  efficiency  results  attained.  Nor  is  it  my  desire  to 
impress  that  idea,  for  the  results  so  far  reached  are  not  entirely 
satisfactory.  More  than  that,  however,  the  endeavor  has  been 
made  within  imposed  limitations  to  make  the  lighting  agreeable 
and  artistic,  and  on  this  point  it  is  hoped  your  interest  and 
sympathies  may  be  enlisted  by  the  description  of  the  work  that 
is  being  done. 

In  order  to  make  the  situation  clear,  it  is  necessary  to  briefly 
describe  the  limitations  under  which  one  has  to  work  in  the 
city  government.  Like  all  artistic  or  utilitarian  things,  illumina- 
tion is.  a  matter  of  money  to  a  certain  degree.  Since  1905,  the 
legislature  has  limited,  properly,  the  prices  to  be  paid  for  illumi- 
nating units,  and  the  Board  of  Estimate  and  Apportionment,  in 
the  budget,  has  limited  from  year  to  year  the  number  of  units  of 
illumination  that  may  be  added  by  setting  the  amounts  of  money 
that  can  be  used  in  each  borough.  There  are  further  limitations 
in  connection  with  opening  streets,  etc.,  which  restrict  the  time 
of  construction,  particularly  in  Manhattan  and  Brooklyn,  so  that 
only  a  certain  amount  of  physical  work  can  be  done  each  year. 
In  consequence,  progress  in  improving  lighting  is  slow  and  con- 
servative to  avoid  frequent  changes,  and  should  be  so,  particularly 
in  the  older  sections  of  the  city.  None  of  you  who  know  the 
illuminating  art  is  willing  to  say  the  last  improvement  has  been 

*  An  address  delivered  t  efore  a  joint  meeting  of  the  Municipal  Art  Society  and  the 
New  York  section  of  the  Illuminating  Engineering  Society,  February  21,  1913. 


200  TRANSACTIONS  I.  E.  S. — PART  II 

made,  or  that  the  last  unit  of  illmuination  has  arrived.  You  also 
know  how  necessary  it  is  not  to  jump  at  conclusions,  or  adopt 
new  units  without  careful  study  and  trial,  and  in  this  term, 
units,  is  included  both  the  light  source,  its  reflectors  and  diffusing 
media.  From  this  you  can  see  that  we  are  limited  to  the 
use  of  standard  tried  lighting  units  and  that  new  ones  cannot  be 
adopted  without  a  working  trial,  for  the  reason  that  the  illumi- 
nating service  purchased  by  the  city  must  not  be  lessened  or 
impaired. 

The  city  contracts  with  public  service  corporations  to  supply 
it  with  certain  illumination  and  illumination  service  on  its  streets. 
That  means  that  the  illumination  is  to  be  given  by  certain  proved 
lighting  appliances,  or,  in  other  words,  proved  units,  and  that 
these  appliances  or  units  are  to  be  kept  in  a  condition  giving  as 
nearly  as  possible  perfect  operation.  The  term  "perfect  opera- 
tion" is  the  definition  of  "good  service."  It  depends  on  many 
elements  and  great  attention  to  detail.  In  1888,  when  first  en- 
gaged in  electrical  work,  a  small  handbook  was  given  me  which 
said  "The  best  electric  system  is  the  man  who  runs  it" — a  crude 
way  of  saying  the  best  illuminating  system  depends  on  the 
efforts  of  the  men  who  run  it.  This  was  in  the  days  when  there 
were  many  different  so-called  systems  of  producing  electricity 
for  illumination  work.  Now,  the  efforts  of  the  men  who  operate 
a  lighting  system  largely  determine  whether  the  householder  or 
the  community  get  good  or  bad  service,  and  the  clumsy  epigram 
quoted  is  just  as  true  now  as  then. 

Good  service  in  electric  and  largely  in  gas  lighting,  depends  on 
the  active  strength  and  reserve  capacity  of  the  central  station  and. 
its  relays,  the  perfection  and  reliability  of  the  distribution  system, 
the  detailed  care  and  attention  given  to  the  most  reliable  lighting 
appliances  obtainable  for  the  system,  and  last,  but  not  least,  the 
discipline,  co-ordination  and  co-operation  enforced  between  the 
above  elements  as  established  and  maintained  by  the  man  in 
charge.  This  man  and  his  loyal  assistants  may  really  be  called  the 
slaves  of  the  lamp,  and  those  of  us  who  have  been  such  slaves, 
know  the  strange  fascination  of  this  work  and  its  demand  for  the 
sacrifice  of  all  else  to  it.  But  that  fascination  and  devotion  mean 
good  service,  and  successful  lighting  service  can  never  be  obtained 


V 


^ 


a  b  c 

Pig.  I.— Gas  lamp  posts:  (a)  Old  gas  iron  type  in  use  about  fifty  years.  Designed  for  open 
gas  flame  o/  about  13  candle-power.  Weigbt  of  lantern  about  14  pounds,  (b)  Early 
mantle  type,  using  lantern  weighing  32  pounds.  Average  candle-power  35  to  40. 
1  c)  Present  mantle  type.  Center  of  illumination  is  9' 6"  above  the  ground.  Average 
candle-power  35  to  40. 


ivWEf 


Fig-  2.— Arc  lamp  posts:     (a)  Bishop's  crook  type.     Design   embodies   suggestions  of  Stanford 
White.     Lamps   trimmed   from   post  or  by  tower  wagon,   there   being   no   lowering  device. 

(b)  Lyre  top  post,  as  used  in  center-parked  plots  on  broad  streets,  such  as  upper  Seventh  avenue. 

(c)  Mast  arm  post,   used  to  avoid  tree   interference.     Used  at  curb  to  overhang  the  street,  in 
some  cases  in  connection  with  the  lyre  top,  as  in  upper  Seventh  avenue  and  upper  Broadway. 


lacombe:    street  lighting  OF   NEW  YORK  201 

without  it;  nor  can  steady,  continuous  illumination— without 
which  any  attempt  towards  agreeable  or  artistic  lighting  is  futile— 
be  obtained. 

Assuming  this  to  be  obtained,  then,  let  it  be  asked  what  choice 
is  offered  in  the  way  of  lighting  appliances  for  service.  It  is 
apparent  on  a  moment's  thought,  that  such  appliances  must  be 
well  tried  and  standard.  New  appliances  cannot  be  used  on  regu- 
lar service.  They  must  be  given  long  service  tests  in  by-ways  and 
unimportant  sections,  often  improved  and  re-tried  until  they  be- 
come reliable  and  of  real  value.  Of  such  available  appliances 
to-day,  there  are  the  arc  lamp  of  standard  design  for  both 
classes  of  current  supply,  and  for  both  multiple  and  series 
systems ;  the  incandescent  lamp,  in  various  intensities,  with 
carbon,  metallized  and  metal  filaments,  for  similar  current  supply 
and  distribution  systems ;  the  mantle  gas  lamp  with  either  vertical 
or  inverted  mantles,  increasing  the  number  of  mantles  for  in- 
creasing illumination;  and,  naphtha-vapor  mantle  lamps,  giving 
the  same  illuminating  effect  as  vertical  mantle  gas  lamps.  In 
addition,  but  not  available  in  all  parts  of  the  city,  are  the  high 
candle-power  arc  lamps  using  a  metallic  electrode,  with  which 
much  can  be  done.  In  Manhattan,  where  the  greatest  congestion 
exists,  long  burning  white  flaming  arc  lamps,  which  afford  still 
higher  illumination  for  general  street  lighting  are  now  being  tried. 
A  limited  number  of  these  lamps  are  being  installed  on  Broad- 
way, above  47th  street  and  on  7th  avenue,  above  34th  street.  This 
list  of  lamps,  then,  comprises  the  lighting  units  which  must  be 
used  at  present. 

This  city  differs  from  others  in  that  low  and  high  tension 
electrical  distribution  are  both  used  on  an  extensive  scale.  In 
Manhattan  the  lighting  is  all  on  low  tension  multiple  circuits, 
both  direct  and  alternating.  In  by  far  the  larger  part  of  Brook- 
lyn and  all  the  other  boroughs  the  street  lighting,  with  a  few 
minor  exceptions,  is  by  means  of  high  tension  direct  or  alter- 
nating current  series  circuits. 

Gas  supply  for  mantle  lighting  is  available  throughout  the 
greater  city,  except  in  a  few  remote  sections.  Naphtha-mantle 
lamps  are  still  used  for  frontier  lighting,  where  neither  gas  nor 
electricity  is  available.     Gas  lighting  was  formerly  used  in  parks. 


202  TRANSACTIONS  I.  E.  S. — PART  II 

but  is  now  being  abandoned  for  obvious  reasons,  in  favor  of  the 
tungsten  incandescent  lamps.  Of  these  various  lighting  units,  in 
the  greater  city,  there  are:  19,180  standard  enclosed  arc 
lamps,  17,991  incandescent  lamps,  78  flaming  lamps,  perma- 
nent and  on  trial ;  44,653  single  mantle  gas  lamps,  28  inverted 
mantle  gas  lamps  on  trial,  1,816  naphtha-vapor  lamps.  In  all  there 
are  83,746  lamps  furnished  by  28  lighting  companies.  Expressed 
empirically,  the  lamps  give  a  horizontal  illumination  approxi- 
mately equal  to  fifteen  million  candles. 

For  the  purpose  of  street  illumination,  these  lamps  are  used 
in  the  following  manner : 

Enclosed  arc  lamps,  reinforced  by  flaming  arc  lamps,  are  in- 
stalled at  points  of  great  congestion,  as  at  Times  Square,  Man- 
hattan. 

Arc  lamps  are  used  on  main  avenues  and  business  streets, 
practically  over  their  whole  length  in  Manhattan,  and  as  far  as 
necessary  in  other  boroughs,  and  then  merged  into  tungsten  in- 
candescent electric  or  gas  mantle  lamps. 

Gas  lamps  are  generally  used  in  the  residence  districts  and  on 
unimportant  streets,  little  used  at  night,  although  of  late,  for 
many  reasons,  the  tungsten  lamp  has  made  great  inroads  on  this 
territory. 

Ten  years  ago  last  month,  when  the  engineering  charge  of  the 
illumination  of  Manhattan  and  the  Bronx  was  put  in  my  hands,  it 
was  found  to  be  inadequate,  unsymmetrical  and  out  of 
date;  the  arc  lamps  on  the  main  avenues  were  insufficient 
and  often  spaced  irregularly,  too  far  apart  and  not  in 
symmetry  in  line  or  in  height,  except  where  standard  fixtures 
were  employed.  Open  flame  gas  lamps  were  in  the  majority, 
with  gas  mantles  and  arc  lamps  often  mixed  in  with  them.  In 
other  words,  there  was  little,  if  any,  system  of  lighting. 

As  soon  as  possible,  this  was  corrected.  First,  prices  were 
materially  reduced  by  continuous  agitation,  and  second,  a  plan  of 
lighting  developed.  All  open  flame  gas  lamps  were  discontinued. 
Arc  lamps  were  practically  doubled  in  number  at  street  and  ave- 
nue intersections  and  symmetrically  arranged  as  often  as 
possible  on  street  house  lines;  so  that  the  illumination  was  made 
as  uniform  as  it  could  be  with  the  funds  allowed.     Mantle  ?as 


(K 


<y> 


Fig-  3- — (a)  Bracket  type  ;  for  use  on  buildings  in  narrow  streets  downtown,  where  a  post  would 
take  up  material  width  of  limited  sidewalk,  (b)  Twin  lamp  posts,  as  used  in  Fifth  avenue, 
(c)  Reverse  scroll  bracket  post,  as  used  on  lower  Seventh  avenue  and  elsewhere  ;  also  used  with 
flame  arc  lamps. 


Fig.  4.— Arc  lamp  posts:  fa)  High  type  ;  used  with  powerful  flame  arc  lamps  in  squares  and  open 
flames,  as  for  example  Times  Square,  t,ong  Acre  Square,  etc.;  height  usually  45  feet,  with  globe 
of  lamp  about  39  feet  from  ground,  (c)  Ornamental  pole,  designed  by  McKim,  Mead  and  White, 
for  flame  arc  lamps  used  around  new  Municipal  Building,  Center  and  Chambers  streets. 


LACOMBE:     STREET    LIGHTING    OE    NEW    YORK  20.3 

lamps  were  put  on  existing  lamp  posts  in  all  residence  streets  ex- 
cepting a  few  quarters  where  arc  lamps  were  already  installed. 
This  arrangement  as  a  whole  has  proved  efficient  and  has  been 
followed  with  improvements  as  they  were  worked  out.  The 
result  may  be  seen  in  the  enlarged  system  of  uniform  distribution 
of  light  sources  in  long  straight  parallel  lines,  exemplified  best, 
first  by  Seventh  avenue  above  Central  Park  and  Fifth  avenue 
from  Washington  Square  to  6oth  street,  and  now  by  almost  all 
the  avenues  in  Manhattan,  from  west  to  east,  running  north  to 
south,  as  well  as  on  the  main  streets  running  east  and  west. 

Late  in  1905,  the  engineering  charge  of  the  illumination  of  the 
greater  city  was  put  in  my  care,  as  chief  engineer,  and  early  in 
1906  such  control  was  formally  assumed.  The  same  general 
system  as  used  in  Manhattan  and  the  Bronx  was  put  into  effect 
in  the  other  boroughs  and  the  bureau  organization  in  these  bor- 
oughs made  to  conform  to  the  simple  system  of  develop- 
ment in  illumination  just  described  in  Manhattan.  It  was 
modified  in  some  ways,  for  in  the  outlying  sections  of 
Brooklyn  and  in  Queens  and  Richmond  the  problem  was 
quite  different.  In  these  boroughs  it  was  the  illumination 
of  small  towns  and  centres,  with  long  connecting  roads 
through  truck  farming  districts  from  centre  to  centre.  These  dis- 
tricts are  lighted  by  means  of  far  flung  alternating  and  direct  cur- 
rent series  systems  of  electrical  distribution  overhead  on  poles, 
with  all  the  limitations  of  such  suburban  systems  and  companies. 
The  same  rule  was  worked  out  in  symmetrical,  continuous  and 
uniform  illumination  with  arc  and  tungsten  lamps,  and  there  are 
now  established  certain  lines  of  well  lighted  roads  from  the  City 
Hall  in  Manhattan,  north  to  Yonkers,  east  to  Nassau  County,  and 
central  Long  Island,  as  well  as  its  north  shore,  southeast  to 
Rockaway,  south  to  Coney  Island,  and  excepting  the  ferry,  to  the 
southernmost  point  of  Staten  Island.  Cross  and  inter-connecting 
roads  through  country  districts  are  also  carefully  illuminated. 
Other  projects  on  this  line  are  now  building  or  planned,  such  as 
the  Boston  Road  and  Pelham  Parkway  and  Park  to  Pelham  and 
the  south  shore  of  Connecticut,  another  route  via  the  Eastern 
Parkway,  Brooklyn,  to  the  Merrick  Road  and  the  south  shore  of 
Loner  Island. 


204  TRANSACTIONS  I.  E.  S. — PART  II 

This  general  system  is  now  established  on  such  basic  lines  that 
it  can  be  developed  consistently  and  without  duplication  of  equip- 
ment expense  using  the  present  lighting  units.  Until  congestion 
spreads  from  Manhattan,  the  lower  portion  of  the  Bronx  and 
Brooklyn,  to  the  outlying  sections,  to  the  extent  that  a  consider- 
able change  from  overhead  line  construction  to  underground  sub- 
ways becomes  necessary,  the  development  need  only  be  to  higher 
units  and  not  a  change  in  systems  or  equipment  to  any  consider- 
able extent. 

A  certain  handicap  will  exist  for  some  years  in  obtaining  the 
best  distribution  of  illumination  in  the  suburbs  on  account  of  the 
overhead  lighting  lines  and  lighting  being  kept  on  one  side  of  the 
street  to  avoid  duplication  of  pole  lines ;  but  this,  in  time,  will  be 
removed. 

In  the  congested  area  of  Manhattan  and  in  the  Bronx,  where 
underground  distribution  exists,  the  locations  of  the  lamps  on  the 
main  avenues  are  secured,  and  to  realize  the  full  perfection  of  the 
present  plan,  only  higher  sources  of  illumination  are  neded,  and 
it  is  to  that  end  the  experiments  with  flaming  arc  lamps  on  Broad- 
way and  Seventh  avenue  are  being  made. 

In  this  development  care  has  been  taken  to  avoid  freak,  or  too 
accentuated  lighting,  on  particular  streets,  as  has  been  the  case  in 
some  cities.  The  attempt  has  always  been  to  keep  to  uniform, 
agreeable  lighting,  avoiding  light  sources  of  high  intrinsic 
brilliancy  and  consequent  glare.  Where  high  intensity  sources  are 
used  they  are  carefully  diffused  or  kept  out  of  the  range  of  vision. 

It  is  all  a  slow,  but  so  far,  sure  progress  of  evolution,  develop- 
ment and  education.  The  results  of  the  original  plan  of  ten 
years  ago  now  show  in  Manhattan,  the  Bronx  and  Richmond,  and 
are  beginning  to  show  in  the  thoroughfares  of  Brooklyn  and 
Queens.  The  mixture  of  lighting  units  has  been  generally  re- 
moved or  is  in  process  of  removal.  In  the  last  ten  years,  the 
illumination  of  the  greater  city,  as  specifically  defined,  has  in- 
creased, under  this  system  from  seven  million  candles  to  fifteen 
million  candles,  or  an  increase  of  118  per  cent. 

It  is  a  peculiar  and  somewhat  odd  fact  about  all  this  develop- 
ment, that  except  in  a  few  isolated  instances,  no  one  in  this  city 
seems  to  realize  the  improvement  made  in  its  lighting  in  later 


y> 


h 

^ 


*       ! 


Fijj.  5  —Tungsten  lamp  posts:  (a)  First  ornamental  tvpe.  French  design:  used  in  Central 
Park  and  on  Riverside  Drive,  (b)  Present  ornamental  type,  designed  by  Henry  Bacon 
for  the  Municipal  Art  Commission  ;  used  in  park  roadways,  (c)  Ornamental  type  with 
diffusing  globe,  designed  bv  Bacon,  used  in  special  places  in  parks,  for  example.  The 
Mall.  Central  Park,  (d)  City  ornamental  type,  now  used  with  500-watt  tungsten  lamp, 
as  shown,  at  Sherman  Square,  59th  street  and  Fifth  avenue.     See  also  Fig.  12. 


Fig.  6. — (a)  Municipal  Art  Society  post ;  designed  by  Mr.  Ciani  and  presented  to  the  city 
by  the  Municipal  Art  Society,  (b)  Astor  post ;  designed  by  Henrik  Wall  in  and  pre- 
sented to  the  city  by  the  Astor  estate.  Two  located  in  front  of  the  Astor  Hotel,  Broad- 
way, (c)  D.  A.  R.  post;  designed  by  Allen  G.  Newman  and  presented  to  the  city  by  the 
Daughters  of  the  American  Revolution  ;  to  be  placed  at  72nd  street  and  Riverside  Drive. 
(d)  Pennsylvania  Station  post;  installed  around  the  Pennsylvania  Station  and  adjacent 
postoffice. 


LACOMBE  :     STREET    LIGHTING    OF    NEW    YORK  205 

years.  Few  newspapers  have  ever  mentioned  it,  and  then  but 
casually;  no  realization  that  a  logical  engineering  scheme  of 
development  is  taking  place  has  been  noticed,  although  it  has 
been  inherently  correct  enough  to  have  been  supported  by  the 
commissioners  of  four  political  administrations  of  the  city  gov- 
ernment. It  is  only  fair  to  say  that  the  technical  journals  and 
societies  have  discussed  it  somewhat,  and  a  number  of  other 
cities  have  noticed  and  studied  the  lighting  and  fixtures  in  use 
here  with  a  view  to  their  use  in  their  own  cities.  In  some  cases 
our  system  has  been  adopted  by  them. 

Even  in  this,  a  sort  of  peculiar  pride  is  taken ;  for  this  city, 


FIFTH  AVENUE 

Present  Installation,  Two  450-Watt  Arc  Lamps  per  Post 
I 1  Over  0.3  Fool-candle      V77\  °'3  "  "•'  F<">,-«ntlle       ;  ^%j  0.1  •  0.03  Fool-candle 


I  0  03  -  0  01    Foot-candle 


Fig.  7. — Diagram  showing  lamp  locations  and  isointensity  curves;  two  450-watt  enclosed  arc 
lamps  per  pole;  Fifth  avenue;  length  of  block  264  feet. 

rushing  on  in  its  career  to  the  goal  of  the  greatest  city  of  the 
world,  has  little  time  to  do  much  but  growl  at  real  and  fancied 
obstructions  to  its  progress.  In  consequence,  to  keep  up  to  the 
extension  of  the  lighting  demanded,  the  lighting  bureau  of  the 
department  has  had  little  time  to  do  anything  but  work,  its  efforts 
being  devoted  to  making  up  for  the  lack  of  development  in  the 
past,  and  hurrying  forward  to  meet  the  present  demands  of  a 
greatly  congested  world  centre.  This  is  really  the  first  address  I 
have  been  able  to  make  on  this  subject. 

In  a  negative  way,  the  work  has  been  noticed ;  for  example. 


206 


TRANSACTIONS  I.  t.  S. — PART  II 


since  the  avenues  have  been  doubled  and  tripled  in  illumination, 
we  have  heard  a  complaint  that  the  side  streets  in  Manhattan  and 
the  Bronx  are  dark,  and  such  complaints  are  always  listened  to 
attentively,  because  they  help  the  engineering  scheme  by  impress- 
ing the  political  and  financial  powers  that  be.  The  side  streets 
are  comparatively  dark,  but  in  1904,  when  the  mantle  gas  lamps 
replaced  the  open  flame  gas  lamps,  these  streets  received  more 
than  three  times  the  old  illumination,  so  one  can  imagine  what 
that  illumination  was.  It  may  be  truthfully  said  that  were  this 
city  thrown  back  in  a  day  to  the  lighting  of  ten  years  ago,  the 
contrast  would  be  so  extreme  there  would  be  a  riot,  almost,  on 


450-Watt  Arc  Lamps 
I  0.3  •  0.1  Foot-candle       V/ZZ\  "■!  ■  0-°3  Foot-candle        Rx>£3  0.03  -  0.01  Foot-candle 


[       ]  Over  03    Foot-candle 

ig.  8. — Diagram  showing  lamp  locations   and  isointensity  curves;  mast  arm  and  lyre  top 
posts;  450-watt  enclosed  arc  lamps;  upper  Seventh  avenue;  length  of  block  264  feet. 

account  of  lack  of  illumination.  In  the  elapsed  time  the  increase 
has  been  slow,  and,  in  consequence,  not  perceptible,  except  from 
year  to  year. 

It  is  desired  to  emphasize  the  fact  that  under  the  engineering 
scheme  all  new  lighting  is  carefully  worked  out  mathematically 
and  geometrically,  on  lines  of  proper  illumination  design.  A 
temporary  equipment  is  then  erected,  movable,  if  desired,  and  the 
final  effect  obtained  by  the  actual  trial  and  observation  on  the 
street.  All  suitable  kinds  of  reflecting  and  diffusing  devices  and 
glassware  are  tried  until  the  best  result  is  reached.  This  example 
is  then  measured  and  plotted  photometrically  and  checked,  or 
corrected,  as  the  case  may  be.  Lighting  is  no  longer  installed 
empirically  or  by  guess  work. 


r 


fth  avenue  twin  arc  lamp  posts  (see  Fig.  3):  location  shown  in  Fig.  7;  night 
photograph. 


Fig.  10.— Upper  Seventh   avenue  ;  mast  arm  and  lyre  top  posts  (see  Fig.  2);  location  as 
shown  in  Fig.  8  ;  night  photograph. 


ii. — Broadway  lighting  with  reverse  scroll  posts  shown  in  Fig 
in  Fig.  13;  night  photograph. 


locations  shown 


Fig.  12. — Plaza  and  Sherman  monument,  59th  street  and  Fifth  avenue,  illuminated  by  flaming 
arc  lamps,  enclosed  arc  and  tungsten  lamps,  night  photograph;  tungsten  lamps  in  the  fore- 
ground and  flaming  arc  lamps  to  the  left  beyond  the  range  of  the  camera. 


lacombe:    street  lighting  OF  NEW  YORK 


207 


The  newer  lamp  posts  used  for  the  improved  lighting  are  also 
designed  with  great  care.  Even  before  the  Municipal  Art  Com- 
mission took  up  this  matter  in  a  systematic  and  effective  manner, 
as  they  did  some  years  ago,  the  New  York  Edison  Company  had 
anticipated  the  demand  for  more  attractive  fixtures  and  has  been, 
and  still  is  pre-eminent  in  this  regard. 

It  is  well  to  state  here,  as  one  of  the  working  conditions,  that 
with  few  exceptions  the  arc  lighting  posts  and  fixtures  belong  to 
the  companies  throughout  the  greater  city,  the  gas  lamp  posts  all 
belong  to  the  city,  as  do  the  tungsten  lamp  posts  in  the  streets  and 


BROADWAY 

Enclosed  Flame  Arc  Lamps 
[ — ~\  Over  0.3  Foot-candle         \7J\  0-3  ■  0.1  Foot-candle  '        |  0.1  -  0.03  Foot-candle 

Fig.  13.— Diagram  showing  lamp  locations  and  isointensity  curves  450- watt 
enclosed  flame  arc  lamps;  Broadway. 

parks  on  underground  service.  The  lanterns  and  lamps  them- 
selves belong  to  the  various  companies.  The  posts  now  supplied 
by  the  lighting  company  are  shown  in  the  accompanying  illustra- 
tions. They  were  first  drawn,  then  life  size  plaster  models  made 
and  revised,  then  the  patterns  developed  and  corrected  until  a 
satisfactory  and  harmonious  result  combining  artistic  effect  with 
engineering  construction  was  attained.  These  posts  were  then 
submitted  to  the  Art  Commission  in  every  case  and  their 
criticisms  embodied  in  the  final  result. 


208  TRANSACTIONS  I.  E.  S. — PART  II 

The  city's  posts  also  go  through  this  process.  In  one  case  the 
Art  Commission  had  designs  drawn  and  paid  for  of  both  the 
lantern  and  posts  for  tungsten  lamps  for  Central  Park.  This 
design  has  been  used  extensively  throughout  the  city  since  and 
has  been  copied  by  other  cities.  The  new  posts  for  side  street 
lighting  will  also  be  submitted  to  them,  when  funds  are  received, 
and  it  is  hoped  that  the  old  ugly  gas  lamp  post  will  shortly  dis- 
appear from  Manhattan. 

It  is  natural  that  the  attempt  to  make  the  lighting  and  fixtures 
artistic  should  begin,  like  the  improved  lighting  system,  in  the 
borough  of  greatest  congestion,  where  it  is  needed  most,  and 
where  the  lighting  is  required  to  be  not  only  useful  but  strong, 
uniform  and  agreeable. 

Manhattan  is  fortunate,  in  the  engineering  sense,  in  being  sup- 
plied with  its  lighting  service  by  a  low  tension  multiple  system, 
which  is  flexible  and  very  adaptable  to  artistic  effects  with  safety. 

The  lighting  current  is  distributed  by  underground  lines,  and 
with  energy  supplied  from  the  greatest  electric  generating  station, 
gives  the  best  obtainable  service. 

It  has  been  my  attempt  so  far  to  give  you,  as  briefly  as  possible, 
the  salient  points  of  the  conditions  under  which  lighting  work  is 
done  in  this  city.  The  endeavor  has  been  made  to  be  non- 
technical. It  is  necessary  to  remind  you  that  the  illustrations 
showing  night  lighting  do  not  at  all  show  the  effect  of  the  same 
lighting  on  the  eye,  so  far  as  the  source  is  concerned,  so  if  you 
will  compare  the  surfaces  illuminated  and  not  the  source,  you  will 
have  a  more  accurate  comparison. 

It  is  my  pleasant  duty  to  acknowledge  how  much  I  am  indebted 
in  accomplishing  this  work  so  far,  first,  to  the  support  of  the 
commissioners  of  the  department;  second,  to  the  collaboration 
and  assistance  of  the  engineering  and  service  side  of  The  New 
York  Edison  Company,  led  by  Vice-president  Lieb,  the  dean  of 
all  central  station  men,  and  to  Mr.  Rhodes,  in  charge  of  the  arc 
lighting  department,  who  has  more  than  done  his  half  of  the  work 
with  me  in  this  development;  third,  acknowledgment  is  also  due 
to  my  own  men,  particularly  to  the  general  inspector  of  street 
lighting,  whose  accuracy  and  great  attention  to  detail  have  done 
much  towards  the  genera    result. 


Fig.  14. — Central  Park  Mall  ;  ornamental  ball  tungsten  lamp  posts  (see  Fig.  5)  niglit 
photograph  ;  60  candle-power  lamps. 


Fig.  15  —  Central  Park  road  lighting;  tungsten  lamp  posts;  (see  Fig.  5-b)  night  photograph; 
60  candle-power  lamps. 


Fig.  16. — Side  street  illumination;  experimental  tungsten  lamp  posts;  locations 
shown  in  Fig.  iS;  night  photograph. 


Fig.  17. — Public  library;  lighting  from  enclosed  arc  lamps  on  the  opposite  side  of  Fifth 
avenue;  approximately  0.1  foot-candle  on  face  of  building;  night  photograph. 


LACOMBE:     STREET    LIGHTING    OF    NEW    YORK 


209 


In  addition  to  the  progress  made,  it  has  also  been  attempted  to 
show  in  this  paper,  in  a  limited  way,  some  of  the  diversified 
problems  of  the  engineer  in  the  lighting  of  this  great  city.  He 
must  proportion  the  lighting  to  the  needs  of  the  various  streets 
or  sections,  and  to  their  importance,  due  to  greater  or  less  use. 
He  must  lay  out  a  system  capable  of  great  increase  without  ex- 
pensive change  of  equipment,  or  the  contracting  companies  will 
object.  He  must  lay  it  out  on  economical  and  efficient  lines  and 
obtain  judicially  fair  prices  within  limits,  or  the  city  administra- 
tion will  object.  He  must  try  to  eliminate  glaring  lighting,  or 
the  Illuminating  Engineering  Society  will  protest  and  he  must 
make  it  agreeable  and  as  artistic  as  possible  or  the  Municipal  Art 
Societv  and  Commission  will  criticise. 


113TH  STREET 

150-Watt  Tungsten  Lamps,  113  Feet  Spacing 
^J  0»ec  0.3  Foot-candle       f\^  0.3-0.1  Foot-candle         |\^|  0.1-0.03  Foot-candle 


0.03-0  01  Foot-candle 


Fig.  iS. — Ump  locations  and  isoititensity  curves  new  tungsten  lamp  posts 
for  cross  streets. 

Agreeable  and  successful  lighting  is  a  combination  of  the  efforts 
of  all  the  different  interests  mentioned.  It  is  not  only  a  matter 
of  artistic  posts,  but  also  artistic  lighting,  if  I  may  use  that  term. 
To  produce  the  best  results,  one  must  have  the  support  of  the  city 
and  the  contracting  company.  The  posts  and  the  lighting  must 
both  be  artistic  and  agreeable.  The  position,  height  and  design 
of  the  supporting  post  have  much  to  do  with  the  effect  of  the  dis- 
tribution of  the  lighting,  as  well  as  its  cost.  It  is  in  these  latter 
details  the  artist  and  engineer  can  well  work  together.  Within 
limitations,  the  artist  must  not  demand  too  extremely,  artistic  a 
design  in  lighting;  it  is  too  costly  except  in  a  few  isolated  in- 


210  TRANSACTIONS  I.  E.  S. — PART  II 

stances.  In  the  general  lighting  of  a  great  city,  the  useful  side 
must  have  as  great  weight  as  the  artistic. 

The  engineer  also  should  not  cling  too  closely  to  the  most 
efficient  and  economical  devices  which  obtain  only  the  greatest 
illumination,  at  the  lowest  cost,  for  the  lighting  must  be  made 
agreeable  to  the  eye.  In  street  lighting,  unless  surrounding  build- 
ings are  to  be  specially  illuminated,  the  useful  rays  are  those  that 
can  be  directed  towards  the  ground.  To  make  this  agreeable,  the 
point  source  of  an  arc  lamp,  for  instance,  must  be  made  into  a 
ball  of  softened  light  by  diffusing  globes  or  shades,  and  this,  so 
far  as  is  possible,  thrown  towards  the  ground  by  either  interior 
or  exterior  reflectors.  This  practise  reduces  efficiency  to  a  cer- 
tain extent  by  the  absorption  of  the  light,  and  demands  either 
closer  spacing  or  more  powerful  sources  of  illumination  at 
greater  heights  from  the  ground ;  it  is,  consequently,  more  expen- 
sive. So  far,  it  has  been  used  mainly  in  Manhattan  and  there, 
principally  on  its  main  avenues. 

In  other  locations,  one  may  see  the  naked  source  of  the  light 
which,  in  turn,  means  glare  that  is  not  as  agreeable,  although 
very  useful  and  less  expensive.  Considerable  success  in  suburban 
sections  of  the  city  has  also  been  attained  by  abandoning  the  in- 
tense arc  unit  and  using  three  to  four  60-70  candle-power 
tungsten  lamps,  and  so  obtaining  better  distribution,  with  less 
glare,  and  at  about  the  same  cost. 

There  is  nothing  new  in  this  plan  of  lighting.  The  ancient 
torch  bearer  always  held  his  torch  as  high  as  possible  to  keep  the 
glare  out  of  one's  eyes  and  to  throw  the  rays  over  as  great  an 
area  as  possible.  The  engineer  uses  his  new  facilities  for  lighting 
in  the  same  way,  puts  them  as  high  as  possible,  and  throws  them 
over  the  greatest  area  he  can.  He  only  adds  to  the  torch  bearer's 
efforts,  the  diffusing  globes  and  the  downward  reflectors. 

Artists,  too,  can  take  advantage  of  the  many  possibilities  of  the 
new  lighting  devices,  both  in  general  and  decorative  illumination. 
Beautiful  color  schemes  are  possible  with  small  colored  lamps  set 
in  mosaic  design,  both  massed  and  in  outlines,  on  walls  of  build- 
ings. As  a  substitute  for  advertising  with  glaring  yellow  flame 
lamps,  the  use  of  mosaic  lighting  would  be  vastly  better  and  just 
as  effective.     Even  general  illumination,  as  given  by  the  high 


lacombe:    street  lighting  OF  NEW  YORK  211 

powered  flame  lamps  can  be  tinted  agreeably  in  contradistinction 
to  dead,  cold,  white  light.  In  other  words,  artistic  lighting  is  as 
possible  as  any  other  form  of  art. 

This  address  would  lose  its  point  if  it  failed  in  suggesting  that 
the  artist  and  engineer  meet  each  other  half  way,  and,  by  com- 
bination, produce  a  joint  result,  obtaining  good,  economical  light- 
ing, consistent  with  artistic  standards.  In  this  way,  lighting  can 
be  designed  which  will  command  complete  approval  from  all 
the  points  of  view  that  may  be  invoked  to  judge  it.  If  such  com- 
bination is  obtained,  then  I  can  say  that  five  years  of  co-operation 
will  make  Manhattan  Island  first,  and  the  greater  city  next,  the 
best  and  most  beautifully  lighted  city  in  the  world. 

DISCUSSION. 

Mr.  Clarence  E.  Clewell  (communicated)  :  The  author 
has  pointed  out  in  a  most  interesting  way  that  the  progress  in 
the  lighting  of  New  York  has  been  made  by  keeping  two  main 
points  in  view,  namely,  first  to  avoid  accentuated  lighting  in  par- 
ticular streets  as  has  been  the  case  in  some  cities  on  the  one 
hand,  and  second,  to  secure  uniform  agreeable  lighting  on  the 
other  hand,  thus  avoiding  light  sources  of  high  intrinsic  bril- 
liancy and  the  consequent  glare.  This  policy  is  highly  com- 
mendable and  the  results  of  an  adherence  to  this  safe  principle 
are  shown  throughout  the  city. 

An  item  of  particular  interest  touched  upon  is  the  definition 
of  good  service,  which  is  given  as  perfect  operation  of  the  light- 
ing appliances.  In  securing  perfect  operation  the  author  has 
aptly  pointed  out  the  necessity  of  not  placing  the  entire  burden 
on  the  lamp  manufacturer,  but  states  in  a  truism  that  "The  best 
electric  system  is  the  man  who  runs  it."  The  full  meaning  of 
this  attitude  is  perhaps  appreciated  most  by  those  who  have  been 
confronted  with  the  operation  side  of  lighting  equipment,  and  a 
good  lamp  coupled  with  the  care  thus  implied  in  its  every  day 
operation,  is  almost  sure  to  result  in  satisfactory  conditions  for 
all  concerned. 

If  in  the  criticism  of  improvements  in  city  street  lighting,  due 
weight  is  placed  on  the  necessity  for  moving  slowly  in  the  adop- 
tion of  untried  apparatus,  such  criticism  will  be  far  less  unreas- 


212  TRANSACTIONS  I.  E.  S. — PART  II 

onable  than  otherwise.  It  is  surprising,  however,  even  with 
the  conservatism  this  imposes,  to  note  the  many  modern  types 
of  lamps  which  are  either  in  regular  service  or  on  trial  in  this 
city.  One  thing  which  is  nearly  always  noticed  where  lighting 
improvements  have  been  effected,  is  the  raising  of  the  standard 
of  illumination  among  even  the  unthinking.  Thus  the  improve- 
ments in  the  lighting  of  the  avenues,  has  lead  to  criticism  of  the 
side  streets  in  Manhattan  and  the  Bronx,  and  this  in  turn  to 
improvements  of  the  side  streets.  This  feature  is  often  a  great 
help  in  extending  higher  and  better  illumination  facilities. 

The  author  has  described  the  methods  of  engineering  con- 
nected with  the  new  schemes  of  lighting.  It  is  a  cause  of  much 
satisfaction  to  know  that  in  the  lighting  of  vast  street  areas,  care- 
ful attention  is  being  given  the  question  of  illumination  design. 
Where  the  artistic  side  is  naturally  given  so  much  weight,  it 
would  be  an  easy  matter  to  place  rather  more  emphasis  on  this 
feature  than  on  the  utility  side. 

The  author  has  pointed  out  a  principle  of  far  reaching  im- 
portance when  he  states  that  the  useful  side  must  have  as  great 
weight  as  the  artistic,  stating  at  the  same  time  that  it  is  equally 
important  not  to  cling  too  closely  to  the  highest  efficiency  of  the 
units  at  the  expense  of  an  agreeable  effect  on  the  eye. 

Throughout  the  whole  address  the  author  has  shown  that 
lighting,  while  not  a  new  question,  is  made  up  of  many  items 
which  are  to-day  looked  at  in  new  ways  and  which  are  solved 
by  new  methods. 

This  is  the  key-note  in  the  progress  of  illumination  at  this 
time,  and  the  fact  that  many  of  the  items  which  concern  the 
final  excellence  of  any  lighting  system  are  often  simple  and 
even  commonplace,  should  not  under  any  circumstances  lead 
us  to  overlook  their  great  importance  to  the  results  obtainable. 

Mr.  J.  W.  CowLES  (communicated)  :  Mr.  Lacombe  sets  forth 
most  interestingly  the  good  results  to  be  accomplished  by  the 
adoption  of  a  broad  and  systematic  scheme  for  street  lighting 
under  the  varying  requirements  which  exist  in  evCry  city,  and 
the  New  York  situation  is  a  striking  example  of  what  can  be 
accomplished  by  close  co-operation  between  the  many  interests 
involved. 


STREET    LIGHTING    OF    NEW    YORK  21 3 

In  many  cities  there  may  be  seen  street  lighting  apparently 
developed  with  the  one  idea  of  illumination  or  utility  in  mind, 
with  practically  no  thought  given  to  the  artistic  features,  which 
are  certainly  of  value  even  though  secondary  to  practical  utility. 
In  other  cases  the  reverse  extremes  are  to  be  noted,  but  in  New 
York  there  is  a  striking  and  pleasing  balance  between  both  the 
useful  and  the  artistic. 

I  believe  that  much  benefit  can  accrue  to  other  municipalities 
and  public  utilities  by  careful  consideration  of  the  points  empha- 
sized in  this  paper. 


214  TRANSACTIONS  I.  Z.  S. — PART  II 

ILLUMINATION  OF  PASSENGER  CARS.* 


BY    J.    Iv.    MINICK. 


Synopsis:  This  paper  presents  a  brief  record  of  the  developments 
in  the  methods  of  passenger  car  lighting  that  have  been  witnessed  since 
1825  when  the  candle  was  the  source  of  light  employed.  Various  types 
of  oil,  gas  and  electric  lamps  which  have  been  in  general  use  are  described 
and  illustrated.  Illumination  readings  and  data  obtained  from  cars  lighted 
with  lamps  of  the  latter  types  are  also  included. 

LIGHT  SOURCES. 

In  their  paper  before  the  American  Society  of  Mechanical 
Engineers  last  winter,  Messrs.  Wood  and  Currie  divided  the 
development  of  passenger  car  lighting  into  four  twenty-five  year 
periods,  beginning  with  the  candle  period  in  1825.  Oil  lamps 
came  into  general  use  about  1850  and  gas  about  1875.  About 
1900  electricity  came  into  use  as  a  means  of  lighting  passenger 
cars  in  steam  train  service,  though  it  had  previously  been  used 
extensively  in  electric  cars. 

Information  concerning  the  early  use  of  candles  is  very  meagre. 
It  is  known,  however,  that  Thomas  Dixon,  the  driver  of  the  first 
passenger  car,  furnished  his  patrons  with  candles.  He  also 
furnished  a  rough  board  table  in  the  center  of  the  car  for  sup- 
porting the  candles.  The  passengers  were  required  to  light  the 
candles  and  tend  their  feeble  flames.  The  board  table  later  gave 
way  to  sockets  attached  to  the  walls,  and  these  were  superseded 
by  fixtures  having  glass  chimneys  to  protect  the  flame  and  a  coil 
spring  in  the  bottom  of  the  socket  to  force  the  candle  upward  as 
it  burned  away,  thus  maintaining  the  flame  at  a  predetermined 
position.  This  type  of  candle  fixture  is  used  extensively  to-day 
as  an  emergency  lamp  in  case  of  failure  of  the  primary  gas  or 
electric  system. 

"Center-lamps,"  with  one  or  more  candles,  came  into  use  dur- 
ing the  latter  part  of  the  candle  period.  Many  labor  saving  con- 
veniences were  developed,  as  for  instance,  an  adjustable  top  to 
hold  the  chimney  in  position  without  the  aid  of  thumb  screws, 

*  A  paper  read  at  a  meeting  of  the  Pittsburgh  section  of  the  Illuminating  Engineer- 
ing Society,  May  16,  1913. 


Fig.  i.  — Center  deck  candle  fixture.     About  1^40. 


Fig.  2.— Center  deck  oil  fixture.     About  iv.o. 


Fig.  3.— Center  deck  gas  fixture.     About  1SS0. 


Fig.  4. — Center  deck  electric  fixture.     About  19 


minick:    illumination  of  passenger  cars  215 

and  brackets,  that  permitted  of  shifting  the  position  of  the  lamp 
both  vertically  and  horizontally. 

While  comparatively  little  has  been  written  concerning  the 
earlier  types  of  oil  lamps  it  is  safe  to  assume  that  they  resembled 
the  candle  lamp  in  general  design.  It  was  comparatively  inex- 
pensive and  quite  convenient  to  remodel  the  candle  fixture  to 
support  an  oil  lamp.  Such  changes  were  very  desirable  as  it  is 
a  very  difficult  matter  in  car  work  to  patch  a  hole  in  the  side  or 
roof  of  a  car  without  showing  the  patch. 

Burners,  wicks,  etc.,  were  adapted  to  the  kind  of  oil  used.  The 
use  of  the  heavier  oils,  such  as  rape  seed  and  Colza  vegetable 
oils,  lead  to  the  development  of  central  draft  burners,  in  which  a 
current  of  air  was  delivered  to  both  sides  of  the  flame  to  produce 
more  rapid  combustion.  The  Argand,  Belgian,  acme  and  student 
lamps  are  representative  types  of  central  draft  burners.  Two 
wicks  feeding  one  flame  was  another  means  of  securing  more 
rapid  combustion,  and  consequently  a  brighter  light. 

Many  of  the  oil  fixtures  were  equipped  with  telescoping  attach- 
ments for  lowering  the  lamps  for  cleaning  and  filling.  Others 
had  removable  oil  reservoirs.  Reflectors  came  into  use  during 
the  oil  period.  In  smoking  cars,  baggage  and  mail  compartments 
cheap  metal  or  mirror  glass  disks  were  placed  back  of  the  lamp  to 
throw  the  light  out  into  the  car.  In  coaches  conical  opal  glass 
shades  were  slipped  over  the  chimney  and  were  supported  by  the 
fixture  arms. 

Coal  gas  was  probably  the  first  kind  of  gas  used  in  lighting  rail- 
way cars.  It  was  secured  from  the  city  gas  mains  and  stored 
in  a  canvas  reservoir,  reinforced  by  wooden  hoops,  in  the  guards 
van.  Iron  pipes,  and  rubber  hose  between  the  cars,  served  to 
connect  the  lamps  to  the  reservoir.  Gasoline  mixed  with  air  was 
very  extensively  used.  Acetylene  gas  was  also  used  to  some  ex- 
tent. Pintsch  gas,  invented  in  1867,  came  into  very  general  use 
on  account  of  its  reliability  and  increased  storage  capacity, 
obtained  by  charging  at  high  pressure. 

A  variety  of  burners  were  used,  the  first- of  which  was  prob- 
ably the  flat  or  "fish  tail"  flame.  There  was  also  a  central  draft 
burner  somewhat  similar  to  the  central  draft  oil  burner.  The 
substitution  of  mantles  greatly  improved  the  quality  of  the  light. 


2l6  TRANSACTIONS  I.  £.  S. — PART  II 

For  reasons  previously  explained  the  earlier  types  of  fixtures 
resembled  the  oil  fixture.  Generally  the  fixtures,  used  prior  to 
about  1905,  were  very  ornamental  in  design  to  correspond  with 
the  interior  finish  of  the  car.  The  introduction  of  the  steel  car 
has  changed  this  condition  and  present  day  fixtures  have  been 
greatly  simplified. 

While  it  had  previously  been  used  in  electric  cars,  the  in- 
candescent lamp  did  not  come  into  general  use  as  a  means  of 
lighting  passenger  cars  in  steam  train  service  until  about  1900. 
Carbon,  metallized  carbon,  tantalum  and  tungsten  filaments  were 
all  used  in  about  the  order  named,  the  latter  type  being  in  gen- 
eral use  to-day.  It  was  the  high  efficiency  of  the  tungsten  lamp 
that  made  electric  car  lighting  possible,  as  the  demand  upon  the 
battery  for  current  was  brought  within  the  necessary  limits  of 
battery  capacity  and  weight. 

The  earlier  electric  fixtures  were  generally  gas  fixtures  re- 
modeled to  take  incandescent  lamps,  many  of  which  were  not 
equipped  with  reflectors.  Such  reflectors  as  were  used  served  as 
dust  collectors,  and  by  thus  attracting  attention,  served  to 
stimulate  the  cleaning  of  the  car. 

The  wide  dissemination  of  knowledge  of  illumination,  and  the 
constant  and  earnest  study  of  the  problems  of  serving  the  travel- 
ing public,  has  resulted  in  better  fixture  designs,  better  distribution 
of  light,  higher  intensities,  and  higher  efficiencies.  Filigree  work 
has  almost  entirely  disappeared.  Simplicity  of  design  has  very 
materially  decreased  initial  costs  and  the  use  of  reflectors  specially 
adapted  to  car  service  has  made  it  possible  to  conceal  the  in- 
candescent filament  without  the  use  of  opal  dipped  or  frosted 
lamps.  It  should  not  be  assumed,  however,  that  the  last  word 
has  been  said  on  this  subject.  There  is  a  wonderful  field  for 
further  development  and  improvement. 

ILLUMINATION. 
While  close  attention  is  now  being  given  to  the  proper  lighting 
of  passenger  cars,  the  chief  effort,  until  within  comparatively 
recent  years,  was  to  reduce  energy  consumption  and  simplify 
methods  of  operation.  Lack  of  attention  to  the  proper  shielding 
of  the  filaments,  the  better  distribution  of  light,  and  the  produc- 
tion of  intensities  sufficient  for  the  comfort  of  passengers,  was 


minick:    illumination  of  passenger  cars 


217 


largely  due  to  lack  of  knowledge  of  this  subject  and  lack  of 
facilities  for  accurately  determining  the  conditions  that  obtained. 
The  development  of  the  candle-foot  photometer  and  other  devices 
has  made  it  possible  to  determine  all  of  these  items  and  wonder- 
ful improvements  have  been  made  in  recent  years. 

The  data  herewith  has  been  selected  from  a  series  of  tests  of 
oil,  gas,  and  electrically  lighted  cars,  conducted  during  the  past 
three  years.  The  cars  were  all  taken  from  regular  service  °.nd 
the  results  are  therefore,  representative  of  service  conditions. 
While  the  dimensions  of  the  cars,  the  spacing  and  height  of  fix- 


CAR 

-    X7^  °ES 

ILLUMINATION 

FOOT-CANDLE  »:«  N6S 

Class                              P.f. 

System                          Oil 

Test                     Stan 

1  -.-. 

1       .64  Ha 

61   lw    .54 

No                                3219 

Fixture           2Lt.Cen.Deck 

Sead-na  Ploir.e 

.i' 

j 

.60  po 

56;2w     .50 

Type                                       *"Ocd 

No  r<»*ures                      4 

Max.  F.C 

.b4 

3 

.64  3a 

52  .''w     .42 

Floor  Area. Sol  Ft.          367 

Spacing         iw**ire>V 

(-1  in.  F.C. 

.24 

s 

•64,4a 

61  4*    .55 

r  n  ;- 

-.   Ihl                              95' 

Aisle  Av.F.t. 

.5.-, 

5 

.61  ,5a 

575k      53 

UoDer  Deck    Liojht6reen 

Rt«:ec*or    CI.Ch^LOpSr.'iM 

•     Seats  A».  F.C. 

.30 

6 

.53  5o 

45  ;w   25 

lower 

LamD           MO-SDua'.  Burner 

NirxJow      ■         ■       » 

.«4 

7 

.4)1    ;7d 

33. 7w    .35 

Above  Belt           •      Oak 

Rating 

Car   Ay.    r.  C- 

.48 

8 

.54  si 

29   9w    .24 

5elew  - 

Lu"-!"S  Jtil'.zed 

114 

.  i  ■  •               -■  c  n  " .  j  1  n 

s-<?   Lvr  •:.-,? 

E-H  ciBncv 

:    '♦•'•!'    [| 


"1 — I — r 


Fig.  5  —Two-light  dual  burner  6xtures— standing. 


tures,  color  of  interior  finish,  etc.,  vary  to  a  slight  degree,  the 
variations  are  not  so  great  that  comparison  of  the  several  types 
of  equipment  cannot  be  made. 

No  special  attention  was  given  to  the  cleaning  of  the  cars  other 
than  to  see  that  the  lamps  were  cleaned  and  in  proper  adjust- 
ment. All  tests  were  conducted  at  night  'and  the  blinds  were 
drawn  to  prevent  the  leakage  of  outside  light.  When  air  draft 
and  temperature  conditions  tended  to  affect  the  value  of  the 
illuminant.   both    standing   and    running   tests    were   made.     All 


2l8 


TRANSACTIONS  I.  E.  S. — PART  II 


CAR 

FIXTURES 

ILLUHIK  ATIOH 

FOOT-CANDLE  READING! 

CIOJ5                                                Pf 

Jvatem.                                 Oil 

Test.                       Runn 

no. 

'1 

1.54 

la     1-27  lw.  US 

Fintunt.           2  Lt  C.ntcr  deck 

Readinc" -plane. 

..36" 

2 

1.29  2a    l-33|2w  l>20 

No.  fixtures                                4 

Maximum   F.  C.  " 

1  44 

J 

I.I  1  pi    M4  Jw   ,.15 

Floor  Area  Sq.rt             400. 

■Spocinq                             I26" 

Minimum    F.  C. 

.62 

4 

14414a     I.i3[4w    1.  is 

FlNI  5H 

Heiqht                                 92" 

Aisle   M.    F.  C. 

J 

1.29  ja     1.25  law    1.02 

Upper  DecK        Olive  Green 

RellectOr    CI    Chitniry    or  dom« 

Aisle  Seats F  C. 

1.   1  J 

t 

86' 6o      .96  c  •*     ■  92 

Above  Belt           Liqht    Oak 

Rot  ma 

Car  Av.  F.  C. 

1     06 

a 

t^L 

Ba     ■  71  law      62 

Below       • 

Lumens  utilised 

4  52 

3eolo                    Red   riosh 

Total     lumens 

Efficiency 

UUUUUULJ14LILJLJLJ , 


! 

#^ 

r=«- 

"S? 

<& 

" f*^ 

l 

Fig.  6. — Two-light  burner  fixtures — running. 


CUR 

FUTURES 

ILLUMINATION 

FOOT  CANDLE  F.EAOINES 

Glass. 

Pf 

Sv3em .                               Oil 

Test,                               Standing 

1        125 

la 

111  llw 

9E 

No. 

BJf.RH  3H0 

Future.               lit.  Center  Deck 

Rearfina  plane 

36' 

it 

113 

?a 

LtOlcW 

105 

!yr*. 

Wood 

Number  futures.                       4 

Ma.imum  fc 

UJS 

3 

■SA 

3a 

LOG 

3W        .99 

Floor  flruisn.ft. 

400 

Spocir.a,.                                  126' 

Minimum  fc 

•SO 

4 

US 

4a 

at 

Aw 

1.0E 

Finish. 

Heiaht.                               qr 

Aisle    av    fc 

LOS 

5 

lei 

5a 

U8 

5w        .97 

UppcrdetK . 

Olivt  qrten 

Reflector  CI  rbimney- op.doim 

Aisle  scats  a*  F  c 

LOS 

b 

■SS 

6a 

•a 

=  w       «5 

L.'vrcrdeci. 

Lamp.                        Acme  burner 

Window  seats  avfc 

■31 

7 

.95 

7a 

«i 

7w      .79 

Above  tin. 

Unfit  Oak 

Rating. 

Carav.fc. 

8 

48 

8a 

•72 

5w       .60 

Below  bttti                      • 

Lumens  per  lamp. 

Lumens  ulililtot 

;su 

Stat). 

Ktd  plush 

Total       lumens. 

fffKiencv 

Fig.  7. — Two-light  burner  fixtures— standing 


minick:    illumination  of  passenger  cars 


219 


nOOT-CAHMJ  PE»D:sSS 


ILLUHIMATIQN 


3   Bwlmi  fgBJBBSfltf  &?Ai! 


jfe    r. •_■<!£ 


'*.-■■'-'* 


f'jr' 


■  IS 

*4.5c. 


Lio)ht  OoTT 


Lamp'     ln»trttal  Artjanot  Bun 


1^*1 


Window  SecTfc  Ay  Ft.C 


JO  Ta      .M7»     .30 


Sfflgjffiaiflim 


E-frTuency  " 


Fig.  8. — Inverted  Argand  burner— running. 


CAR 

FUTURES 

ILLUMINATION 

FOOT    CANOLF.  RFAflNtS 

Uau 

S>iH"     (.i.VL-r*(r    &JMlmi 

Tert                                 Standing 

1          1>» 

U          Lit 

V        .» 

No                               WJ 

ll»H  51U 

Futures            1  Lt    CcnTtr  deck 

Read^q  plan!                                 3e" 

t           •» 

l«         .67 

1»        7* 

Wood 

No  f.ilom                                          5 

M»»       I.e..                                    155 

3            .74 

3.         .10 

3«         t5 

F't.f  0r(|  1|  ft 

»! 

Sp.t.na                                             »•' 

Hin         ft                                          -3» 

4        n 

««        -7« 

tw         M 

nnuh 

Hf.qh!                                                  "W 

.     .  .       ..   •  i                                     •» 

S            115 

So           US 

5«        M 

.    -<    #.  ■ 

Ol'wt  qrifn 

Rldfdor                                    CI.  Ajobt 

A    |1|      ^.^»  .».  <    <                               .10 

i             -78 

S«         .71 

6»       M 

LI«H*C4 

Lamp  imtrtid  Ara*i«*  Vomer 

*T«o™  stall  av  f.c             .71 

7           •  »» 

7a           .» 

7"       ■» 

«t..l  [<  • 

r-  :*• 

Rot*, 

Cor      «  re                                  .78 

v  ....  • 

Lomtns  utililM                       116 

'.J, 

St.*  0  jiS 

'.-3'         .-(-A 

tfF,c;tntr 

UUUUUUUUUUUU ^ 

riifjiTfi  ilifnf M 


Fig.  9.  — One-light  inverted  Argand  burner— standing. 


220 


TRANSACTIONS  I.  E.  S. — PART  II 


running  tests  were  made  while  the  cars  were  running  at  a  uni- 
form speed  of  about  40  miles  per  hour.  So  far  as  possible,  all 
equipment  was  adjusted  to  operate  at  the  manufacturers  rating 
and  where  such  adjustment  was  not  possible  proper  correction 
has  been  allowed.  In  the  case  of  running  tests,  lamps  were  ad- 
justed for  running  conditions  and  no  change  in  adjustment  was 
made  for  standing  tests.  Foot-candle  readings  were  taken  on  the 
horizontal  plane  36  (0.914  m.)  inches  above  the  floor. 

Figure  5  shows  a  standing  test  of  dual  burner  oil  lamps. 
This  type  of  fixture  consists  of  two  lamps  spaced  on  about  16- 
inch  (40.64  cm.)  centers  cross-wise  of  the  car.     Each  lamp  has 


Sins. 


Floor  Area, Set  Ft 


5a.  1 

HE 


IhiSH., 


Upper  Deck-    Light  Oli/eSreon 


ZSaKI 


Fixtures        Z  it.  Center  Deck 


gp?c,lnq 


Height 


^effector       CI.GI.'iVo  o>6"  'in 


tamp    "_  Ihy,  Am«nd  Burner. 


•   ■,.■   ■ 


*.<-P.. 


kymsm  per  Lamp   _Js 

T°tal  Wnena .    r  ifi 


ILLUMINATION 


Reading  Plane 


.-.'ar-ling   I  1.02 


Alsl»  Av.  F.C. 


BEES 


\y.r!\^V 'VriViz'e.'d' 


S21 


msney 


v  ; 


FOOT-CANDLE  READIN65 


WIS 

5i  I  I     I     Mi 
<"* 

a 

z3 

o 

►"» 

o' 

o 

k.       


Fig.  10. — Two-light  inverted  Argand  burner— standing. 

two  wicks  feeding  a  single  flame.  Oil  is  fed  through  tubes  from 
a  common  central  reservoir.  Each  lamp  was  equipped  with  clear 
chimney  and  conical  opal  glass  reflectors.  Burners  were  adjusted 
so  that  the  center  line  of  the  flame  was  at  right  angles  to  the  axis 
of  the  car. 

Figures  6  and  7  show  running  and  standing  tests  respectively 
with  Acme  burner  oil  lamps.  This  is  a  central  draft  type  of 
burner.  Each  lamp  was  equipped  with  a  clear  glass  chimney,  and 
an  opal  reflector.  The  illumination  was  pleasing  and  sufficient  for 
newspaper  reading.  The  light  interior  finish  added  to  the  appear- 
ance of  the  car. 


minick:    illumination  of  passenglr  cars 


221 


r00T-C*flPU  PUP1HG5 


1  Li  Center  Deck 


l%h 


lofci  n 


rmi3 


Upper  Deck  Pea  Grgg  n 


Height. 


rroiV_i  Gi^rVv^v 


APo<«  Pel'  l.iq>it  Oak 


pjas: 


f-o-vnfluih 


tr— U  UUUUUU4L444L4LJU, 


ijiiy'ftiitiiifflifin'lF^ 


5m  i     I    || 

4     L   —  . 

J 


Fig.  ii.— One-light  single  mantle — running. 


CIS 

FUTU 

RES 

ILLUHINATI 

□  N 

.    FOOr-CAHOLE  KEADlNtS 

On 

SysUi 

Pints*  C«s 

Tut 

btandinq 

l          V  |Ui        tS8 

lw         1* 

Ho 

PR  0  •        X>9 

Fotures 

I  Lt  Center  dec*. 

Reoctna  plane 

36- 

2           220 

20          2J7 

tw      tS 

Trr» 

No  Futures 

rtarnnumfc 

7.9S 

3           225 

3a          2J0 

!w        187 

floOfa'ca  soft 

i« 

Spacma,                                         1061 

»                                       122 

4           2JI 

'  i            yi 

•»    r 

Finish 

Heiqht 

95 

Aisle  or  fc 

!2« 

s       m 

ia        m 

SW         « 

ppvrAd 

reaqreei 

Reflector 

FmtulGMeOI   ! 

Aisle  scats  or  fe 

22Z 

6             I9t 

(a        ■..«.« 

6w       1>S 

mi  fled 

Lamp 

Laroe  Mantle  3M4 

window  seats  a*  fc 

04 

I         It: 

7a         Vl« 

7*     5s 

AW«e  bill 

UqMOpk 

Reettiq 

»e  s  n  s  c  p 

Carav  fc 

2(4 

8             131 

Sa         01 

!W         112 

Below  belt 

Lumens  per  Lamp 

Lumens  utilities 

Joe 

Seats 

Brown  Hull 

Total  Ihiis 

3MS 

Efficiency 

.Us* 

r-^JUUUUUUUUUUUUM 

m  m  mrp  m  rfi  m  ffim 


nl 


r 

1 

Fig.  12— One-light  single  mantle— standing. 


222 


TRANSACTIONS  I.  E,.  S. — PART  II 


ILLUMINATION 


SJS55 


FOOT- CANDLE  READINGS" 


'ypf 


fixture       I  Lt  Center  Deck. 


No.Fixturw 


Reqd  ir\c^  Plane 


K.,,    F.C 


g      2.«2l2»    2.7^|gw  1.76 


rc 


27SJX1    2.4l|wi 


'■!••/■ 


t'Beit  UnjjrSTtfSJliijhrOS^ 


L  nmp  50  w  £:>    luo.plen 


fi,'4>Le  ■Jeais  a*  fC         £.35 


5       2:76:Ja    2,5J,i>'  '    7^" 
9     JL.4%1<\     2.,'',". 


y  u  lu  u  u  u  u  u.  u  u  u  u  u  u  g  u  y  lj  Lj  ua  u  ii__j 


Fig.  13. — One-light  electric  fixture— standing. 


a 


CAR 

FIXTURES 

ILLUMINATION 

FOOT  CANDLE  READINGS 

CI065                                               P-70 

ElectnQ 

'est                                 Stondinq 

la      1>3 

iw      W 

No. 

Fixture 

'  t   Cc-'ter  Beck 

kcaui.nq  plane                               36 

Z        333 

2a      Ml 

lv|     1.5) 

Type                                S'ee! 

No  FiAtur 

5                                  10 

Mat.    f.c                                  ).8t 

in      \3C 

>tv     .36 

Floor  Area,  td.  Ft.              538 

Scccinfl 

Min.    f-c.                             Ill 

4        Ml 

4a     339 

4«    w 

FINISH 

Heiaht 

Aisle    o».U.                          3.S7 

5       W5 

5q      V5 

'*   :m 

Reflector                      \UU  S.F. 

).'.',■  stitsa.U                           126 

£      lis 

"5a"    lib 

*.-.    nJ 

Lamp 

; .    3> 

if.   ;»o 

Car    av.  t.c.                                307 

3       *; 

s  a    3;: 

Bn.  «S 

Below     »            Dark  Gr.cn 

lomo              409 

Lumens  utilized                1652 

1       3.31 

9  a     Ml 

i»  ni 

Seats                 Green  Flush 

Wlumei 

•;                     40TO 

EttlClMlcY                                 «,«% 

U  111  U  U  U  U  U  U  U  U  U  LTD  Lj  D  U  U  jJD  U 


Fig.  14.— One-light  electric  fixture — standing 


I 


minick:    illumination  of  passenger  cars  223 

Figures  8  and  9  show  running  and  standing  tests  re- 
spectively of  central  draft  types  of  carburetted  gasoline  equip- 
ment. The  burner  is  inverted  in  a  clear  glass  bowl.  Air  from 
the  train  system  is  passed  through  a  spiral  tube  containing  a  wick- 
soaked  in  gasoline,  and  the  resultant  mixture  passes  to  the  lamp 
to  be  burned.  The  carburetter  is  placed  on  the  top  of  the  car 
above  the  lamp  so  that  the  heat  of  the  air  raising  from  the  lamp 
will  heat  the  incoming  gas.  This  type  of  lamp  is  very  susceptible 
to  changes  of  temperature  and  draft,  and  therefore  requires  con- 
stant attention  to  prevent  smoking,  especially  during  stops  at 
stations. 

Figure  10  shows  the  results  obtained  by  the  use  of  Pintsch 
gas  with  inverted  Argand  burner  lamps. 

Figures  11  and  12  show  running  and  standing  tests  re- 
spectively of  single  mantle  Pintsch  gas  lamps.  Gas  is  carried  in 
tanks  under  the  car  at  a  pressure  of  about  150  to  160  pounds  per 
square  inch  (6.45  sq.  cm.)  and  is  reduced  at  the  lamp  to  about 
two  pounds  per  square  inch  (6.45  sq.  cm.)  "With  temperatures 
below  20  deg.  F.,  some  of  the  hydro-carbons  are  precipitated  with 
a  consequent  reduction  in  illumination.  It  will  be  noted  that 
there  is  practically  no  difference  in  illumination  between  running 
and  standing  conditions. 

Figures  13  and  14  show  the  illumination  that  was  obtained 
in  steel  cars  equipped  with  50-watt.  60-volt  tungsten  filament 
lamps,  the  first  with  flat  prismatic  reflectors  and  opal  dipped 
lamps  and  the  latter  with  satin  finish  prismatic  bowl  type 
reflectors  with  clear  lamps. 

CONCLUSIONS. 

From  these  tests  some  idea  of  the  comparative  values  of  the 
several  types  of  car  lighting  units  may  be  had.  Similar  tests  of 
cars  varying  only  in  interior  color  and  finish  will  show  changes  in 
efficiency  of  100  per  cent,  or  more,  and  other  tests  of  direct,  semi- 
indirect  and  indirect  fixtures  will  show  considerable  variation  in 
current  consumption.  The  standard  car  lighting  battery  is  none 
too  large  from  a  capacity  stand-point,  while  from  a  weight  stand- 
point it  is  now  as  large  as  it  can  be  made  for  convenience  in 
handling.     Changes  in  lamp  efficiencies  will  be  of  little  value  until 


224  TRANSACTIONS  I.  E.  S. — PART  II 

they  can  be  increased  sufficiently  to  either  reduce  the  weight  or 
increase  the  hours  of  battery  service  from  25  to  50  per  cent. 
Fixtures,  reflectors  and  color  of  interior  finish  should  receive 
further  attention  with  a  view  of  reducing  maintenance  and 
operating  costs  and  providing  better  distribution  of  light  at  higher 
efficiencies.  Comparative  tests  should  be  made  with  side  light 
semi-indirect  and  indirect  fixtures  to  determine  their  value  not 
only  from  a  standpoint  of  illumination,  but  from  the  more  im- 
portant stand-point  of  current  consumption. 

DISCUSSION. 

Mr.  P.  S.  Millar:  I  have  enjoyed  Mr.  Minick's  talk  very 
much.  His  sketch  of  the  history  of  the  development  of  railway 
cars  is  very  interesting,  while  that  part  of  it  which  pertains  to 
car  lighting,  and  which  is  included  in  the  printed  paper,  sets 
forth  a  record  of  the  improvement  in  this  phase  of  illumination 
which  heretofore  has  been  lacking  in  our  Transactions. 

In  intensity  of  illumination  the  following  is  recorded  in  the 
way  of  improvement: 

Oil — flat  wick 0.57  foot-candle 

Oil — center  draft;  gasoline  and  Pintsch  gas.  .0.86  to  1.04 

Pintsch  mantle  1.99 

Electric    2.60 

But  it  is  not  alone  in  intensity  that  improvement  has  been 
effected.  The  illustrations  show  that  in  the  latest  installations 
with  electric  lamps,  the  filaments  are  shielded  from  ordinary  view 
and  depolished  reflectors  are  used.  These  installations  show  the 
result  of  attention  to  car  lighting  which  is  very  gratifying  to  all 
of  us  who  have  to  use  the  trains  and  who  have  in  the  past 
suffered  not  only  from  inadequate  lighting  but  also  from  bajl 
lighting. 

Recently  I  made  a  hurried  trip  from  New  ork  to  Boston  and 
return.  My  seat  was  in  a  car  illuminated  by  a  large  number  of 
small  tantalum  lamps.  The  decoration  of  the  car  was  dark  so 
that  not  only  the  ordinary  glare  from  the  lamps  led  to  discom- 
fort, but  this  was  enhanced  by  the  great  contrast  of  the  lamps 
against  the  dark  background.  On  the  return  journey  I  was 
accompanied  by  a  member  of  this  Society  who  is  prejudiced 
against  indirect  lighting.  It  happened  that  we  traveled  in  one 
of  the  new  trains  on  the  New  Haven  road,  which  is  equipped 


ILLUMINATION  OF  PASSENGER  CABS  225 

throughout  with  indirect  lighting  fixtures.  After  a  casual  obser- 
vation, my  companion  commented  adversely  upon  the  illumina- 
tion, speaking  of  inadequacy  of  light,  cheerless  appearance,  etc. 
After  five  hours  in  the  train,  however,  both  of  us  were  ready  to 
say  that,  ocularly  speaking,  we  had  never  had  a  more  comfort- 
able railway  journey.  The  single  feature  of  concealed  light 
sources  which  characterizes  indirect  lighting  was  a  most  pleas- 
urable element  of  lighting  of  a  class  in  which  exposure  of  light 
sources  is  all  too  often  the  most  prominent  feature. 

All  things  taken  into  consideration,  I  think  we  may  well  feel 
gratified  at  the  improvements  in  car  lighting  which  are  being 
effected  by  the  leading  railroads,  of  which  the  installation  just 
referred  to  is  one  example. 

Mr.  A.  C.  Cotton  :  There  are  three  general  methods  that 
are  or  have  been  in  use  to  produce  energy  for  illuminating  elec- 
trically lighted  cars.  The  first  method  was  by  the  use  of  a 
straight  storage  battery  which  had  its  difficulties  on  account  of 
the  lamps  sometimes  being  put  on  the  battery  while  it  was  being 
charged,  the  lamps  thus  receiving  a  much  higher  voltage  with 
a  consequent  shortening  of  the  life  of  the  lamps.  When,  there- 
fore, the  voltage  dropped  off  to  1.8  volts  per  cell,  the  illumina- 
tion was  rather  poor.  Another  method  was  by  the  use  of  a 
generator  driven  by  a  steam  engine  or  turbine  placed  in  the 
baggage  car  next  to  the  locomotive,  steam  being  furnished  to 
these  units  through  a  hose  connection  from  the.  locomotive.  In 
some  cases  the  turbine  was  placed  directly  on  the  locomotive. 
At  times  when  there  was  a  failure  of  steam  on  the  locomotive, 
the  engineer  either  throttled  the  steam  to  the  generator  unit,  or 
turned  it  off  altogether,  which  caused  a  failure  of  the  electric 
iights  on  the  entire  train.  One  great  trouble  with  this  system 
was  that  in  the  event  of  the  necessity  of  taking  any  one  of  the 
electrically  equipped  cars  from  the  train,  and  substituting  there- 
for either  an  oil  or  gas  lighted  car,  all  electrical  lighted  cars 
back  of  this  car  were  in  darkness.  This  caused  considerable 
discomfort  to  the  passengers  and  the  railroad  companies  were 
severely  criticised.  On  account  of  this  and  other  difficulties, 
the  head-end  systems,  whereby  the  cars  were  illuminated  from 
a  common  train-line  without  the  use  of  storage  batteries,  have 
been  quite  generally  discarded. 


226  TRANSACTIONS  I.  E.  S. — PART  II 

The  system  quite  generally  used  to-day  is  that  known  as  the 
axle  generator  type,  wherein  each  car  is  equipped  with  a  gene- 
rator suspended  from  the  truck  beneath  the  car,  and  driven  by 
a  pulley  fastened  to  one  of  the  axles.  Each  car  is  also  equipped 
with  a  storage  battery,  generator  regulator,  for  controlling  the 
generator,  and  a  lamp  regulator,  for  controlling  the  voltage  of 
the  lamps.  There  are  several  different  types  of  axle  generator 
systems  in  use  to-day,  both  foreign  and  domestic,  but  the  ones 
we  are  principally  interested  in,  are  those  manufactured  in  this 
country,  as  the  foreign  types  have  gained  very  slight  foothold 
up  to  the  present  time.  The  generator  in  this  system  is  equipped 
with  a  pole-changer,  so  that  the  polarity  of  the  wiring  is  always 
the  same,  irrespective  of  the  direction  of  the  movement  of  the 
train.  The  generator  regulator  governs  the  point  at  which  the 
generator  is  thrown  in  on  the  battery  and  also  governs  the  output 
of  the  generator  by  changing  the  strength  of  the  field  of  the 
same.  The  lamp  regulator,  which  may  be  either  motor  operated 
or  magnetically  operated,  acts  so  as  to  keep  the  voltage  of  the 
lamps  constant,  irrespective  of  the  voltage  of  the  generator  or 
battery.  By  means  of  this  system,  the  storage  batteries  on  the 
cars  are  automatically  charged  while  the  train  is  in  motion,  the 
generator  supplying  current  to  either  the  battery  or  lamps,  or 
both,  while,  when  the  train  is  standing,  the  current  is  supplied 
to  the  lamps  from  the  battery. 

Mr.  J.  L,.  Minick  (In  reply)  :  The  use  of  electricity,  as  a 
means  of  lighting  passenger  cars  on  the  Pennsylvania  Railroad, 
was  brought  about  largely  by  reason  of  the  improvements  in 
the  vicinity  of  New  York  City.  On  account  of  the  large  amount 
of  trackage  in  tunnels  under  the  North  and  East  Rivers  and 
Manhattan  Island,  it  was  thought  wise  to  use  only  such  equip- 
ment as  contained  no  inflammable  materials  of  any  kind,  conse- 
quently electricity  became  the  agent  for  supplying  light.  As 
practically  all  of  the  steel  equipment  is  likely  to  enter  New  York, 
electric  light  has  become  the  prevailing  system. 

The  length  of  run  on  a  single  battery  charge,  will  of  course, 
depend  upon  the  demand  for  energy  during  the  run.  The  bat- 
teries on  the  cars  of.  the  Pennsylvania  Railroad  have  a  rated 
capacity  of  300  ampere-hours.  Some  of  the  express  cars  have 
a  maximum  current  demand  as  low  as  2.4  amperes,  while  certain 


1 1.]. I'M  I  NATION  OF    PASSENGER  CARS  227 

dining  cars  may  exceed  45  amperes.  So  that  the  length  of  run 
may  vary  from  125  hours  to  about  6.5  hours  of  lighting.  Coaches 
with  63-volt  equipment  have  a  current  demand  of  about  10  am- 
peres. Coaches  have  been  run  from  New  York  City  to  St. 
Louis  on  one  charge.  I  am  not  prepared  to  give  the  length  of 
run  possible  on  a  full  gas  charge. 

Referring  to  Mr.  Millar's  remarks,  I  wish  to  say  that  from 
the  standpoint  of  illumination  alone,  I  am  very  favorably  in- 
clined towards  some  form  of  indirect  lighting.  In  car  work, 
however,  there  are  many  considerations  in  addition  to  that  of 
proper  illumination.  Standards  have  been  fixed  at  an  average 
intensity  of  three  foot-candles  for  coaches  and  about  five  foot- 
candles  for  dining  cars.  The  size  and  weight  of  the  present 
300  ampere-hour  battery  is  as  great  as  it  can  possibly  be  made 
for  convenience  in  handling,  while  from  a  current  capacity  stand- 
point it  is  none  too  large.  If  semi-indirect  lighting  be  used  the 
current  demand  will  probably  be  increased  by  100  per  cent,  for 
the  same  intensity  and  if  totally  indirect  be  used  it  may  be  in- 
creased by  several  hundred  per  cent.  If  it  be  attempted  to  main- 
tain equal  brilliancy  in  the  car  the  current  demand  will  again 
be  increased  several  times.  Under  present  conditions  I  do  not 
believe  it  possible  to  provide  either  semi-indirect  or  totally  in- 
direct lighting  at  a  reasonable  demand  for  energy. 

When  berth  lamps  were  first  used  they  were  a  decided  im- 
provement over  the  then  existing  conditions.  I  understand  that 
an  effort  is  now  being  made  to  entirely  conceal  the  lamp  but 
the  space  available  in  existing  cars  is  probably  so  limited  that 
changes  for  the  better  cannot  be  made  conveniently.  In  the 
lighting  of  trolley  cars  it  is  necessary  to  place  several  lamps 
in  series  across  the  circuit  on  account  of  the  high  voltage  used. 
As  the  failure  of  any  lamp  in  this  series  puts  the  entire  series  out 
of  commission  it  is  necessary  to  provide  several  series  circuits 
for  a  single  car.  In  a  small  car  this  means  a  large  number  of 
small  lamps.  If  the  total  number  of  lamps  could  be  reduced  I 
have  no  doubt  more  consideration  would  be  given  to  correct 
lighting. 

Mr.  L.  C.  Porter:  There  are  two  sentences  on  the  seventh 
page  of  Mr.  Minick's  paper,  to  which  I  would  like  to  call  atten- 
tion.   They  are  as  follows :    "The  illumination  was  pleasing  and 


228  TRANSACTIONS  I.   E.   S. PART  II 

sufficient  for  newspaper  reading.  The  light  interior  finish  added 
to  the  appearance  of  the  car."  It  seems  to  me  that  there  should 
be  a  great  deal  more  attention  paid  to  the  esthetic  effects  of  the 
interior  finish  of  cars,  than  is  at  present.  If  a  car  is  finished 
entirely  in  dark  color,  no  amount  of  light  will  make  it  appear 
bright  and  cheerful. 

Photometer  tests,  while  useful  in  comparing  the  actual  effi- 
ciency of  utilization  of  different  systems,  do  not  tell  the  complete 
story,  and  should  be  supplemented  by  personal  observation. 

Mr.  Minick  has  asked  for  a  discussion  of  the  plane  on  which 
illumination  measurements  should  be  made  in  railway  cars,  call- 
ing attention  to  the  fact  that  some  people  are  making  them  on  a 
45  deg.  plane,  which  is  the  plane  in  which  a  reader  would  natu- 
rally hold  a  paper,  while  others  are  making  them  on  the  horizon- 
tal plane.  I  believe  both  planes  should  be  used.  It  is  necessary 
to  find  the  average  illumination  on  the  horizontal  plane,  in  order 
to  calculate  the  effective  lumens  and  thus  determine  the  efficiency 
of  utilization  of  the  lighting  system,  interior  finish  of  the  car,  etc. 
Measuring  on  the  horizontal  plane,  however,  will  not  take  into 
account  shadow  effects. 

With  high-power  units  and  wide  spacing,  a  passenger  at  the 
end  of  the  car,  seated  with  the  first  illuminant  a  considerable  dis- 
tance in  back  of  him,  might  receive  good  light  on  his  paper,  while 
on  the  other  hand,  if  he  were  facing  the  lighting  unit,  the  side 
of  the  paper  towards  him  would  be  in  shadow. 

I  believe  that  at  each  station  readings  should  be  taken  on  three 
planes,  namely,  the  horizontal,  the  45  deg.  towards  one  end  of 
the  car,  and  the  45  deg.  plane  towards  the  other.  Multiplying 
the  horizontal  reading  by  the  ratio  of  the  two  45  deg.  planes, 
always  dividing  the  low  reading  by  the  high,  will  give  a  figure 
which  will  show  more  nearly  the  effectiveness  of  the  illumination. 
For  example,  suppose  the  intensities  on  these  three  planes  with 
two  different  lighting  systems  were  as  follows : 

Foot-candles  Foot-candles  Foot-candles 

Horizontal  plane  45  deg.  rear  45  deg.  front 

Sj'stem  A 3  3. 20                        2.56 

System  B 3  3.20                        1.60 

2. 56 
Then  from  system  A"  we   have  3  X  — —  =  2.4,  and  from  B  we 

3.20 

get  3  X   — —  =  1.5.     Clearly  system  A  is  the  better  system. 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 


Published  monthly,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIII 


JUNE,  1913 


No.  6 


Council  Notes. 

A  regular  meeting  of  the  Council  was 
held  in  the  general  offices  of  the  society, 
29  West  39th  Street,  New  York,  June 
13,  1913.  Those  present  were:  Preston 
S.  Millar,  president;  Charles  0.  Bond, 
George  S.  Barrows,  L.  B.  Marks,  Nor- 
man Macbeth,  Joseph  D.  Israel,  general 
secretary,  W.  J.  Serrill  and  George  H. 
Stickney   by   an   invitation. 

A  monthly  report  on  the  finances  and 
membership  of  the  society  was  re- 
ceived from  the  general  secretary. 

Upon  recommendation  of  the  Finance 
Committee,  bills  aggregating  $1,833.99 
were  authorized  paid. 

Twelve  applications  for  membership 
and  four  resignations  were  accepted. 
Counting  these  changes,  the  member- 
ship was  said  to  total  1,373  members. 

Reports  of  progress  were  received 
from  the  following  committees :  Recip- 
rocal Relations  with  Other  Societies, 
Collegiate  Education,  Advertising,  No- 
menclature and  Standards,  Progress, 
Research  and  Glare. 

A  report  was  received  from  the 
Committee  on  Tellers  giving  the  results 
of  the  election  of  officers  for  the 
society  and  the  several  sections.  The 
names  of  the  officers  elected  appear 
elsewhere  in  this  issue. 

A  report  of  progress  on  the  work 
of    the    Philadelphia    Section    was    re- 


ceived from  Vice-President  W.  J. 
Serrill. 

Two  applicants,  The  Consolidated 
Gas,  Electric  Light  &  Power  Company 
of  Baltimore  and  the  Welsbach  Com- 
pany, were  elected  sustaining  members. 

Additional  appointments  by  the  presi- 
dent to  various  committees  were  ap- 
proved. 

The  general  secretary  was  directed  to 
prepare  a  report  on  the  work  of  the 
Council  during  the  present  year  to  be 
submitted  to  the  membership  of  the 
society. 

The  Executive  Committee  was  em- 
powered to  act  for  the  Council  during 
the  summer  months. 

A  communication  from  the  Heights 
of  Buildings  Committee  of  the  City  of 
New  York  inviting  the  society  to  co- 
operate with  the  committee  in  its  work 
was  read.  The  president  was  directed 
to  acknowledge  the  communication  and 
to  state  that  the  society  will  be  glad  to 
co-operate  with  the  committee  as  far 
as  possible. 


New  Members. 

The  following  applicants  were  elected 
members  of  the  society  at  a  meeting  of 
the  Council,  June  13,  1913: 
Butler,  Henry  Emanuel 

Asst.  in  Illuminating  Engineering 
Laboratory,  General  Electric  Com- 
pany, Schenectady.  N.  Y. 


TRANSACTIONS   I.    B.    S. — PART   I 


English,  J.  C. 

President,  J.  C.  English  Company, 
128  Park  Street,  Portland,  Oregon. 

Freemen,  E.  H. 

Professor  of  Electrical  Engineering, 
Armour  Institute  of  Technology, 
33rd  Street  &  Armour  Avenue,  Chi- 
cago,  111. 

Johnson,  Otis  L. 

Illuminating  Engineer,  Benjamin 
Electric  Mfg.  Company,  120  So. 
Sangamon  Street,  Chicago,  111. 

Klingman,  A.  M. 

Asst.  Commercial  Engineer,  Na- 
tional Quality  Lamp  Division  of 
General  Electric  Company,  Nela 
Park,  Cleveland,  O. 

Langan,  Joseph 

Assistant  Secretary,  Illuminating 
Engineering  Society,  29  West  39th 
Street,  New  York,  N.  Y. 

Roland,  E.  U. 

Headlight  Field  Man,  Remy  Electric 
Company,  Anderson,  Ind. 

Schott,  Albert 

Electrician,  McCreery  &  Company, 
6th  Avenue  &  Wood  Street,  Pitts- 
burgh, Pa. 

Simpson,  Richard  E. 

Asst.  in  Illuminating  Engineering 
Laboratory,  General  Electric  Com- 
pany, Schenectady,  N.  Y. 

Strang,  Perry  S. 

Inspector,  National  X-Ray  Reflector 
Company,  11 19  West  Jackson  Boule- 
vard, Chicago,  111. 

Wangersheim,  E.  A. 

President,  General  Lighting  Fixture 
Company,  28  West  Lake  Street, 
Chicago,  111. 

Wood,  Douglass 

Illuminating  Engineer,  Bryan-Marsh 
Electric  Works,  431  So.  Dearborn 
Street,  Chicago,  111. 


Sustaining  Members. 

At  a  meeting  of  the  Council,  June  13, 
1913,  the  Consolidated  Gas,  Electric 
Light  and  Power  Company  of  Baltimore 
and  the  Welsbach  Company  were  elected 
sustaining  members  of  the  society. 


Section  Notes. 

CHICAGO    SECTION. 

At  a  meeting  of  the  Chicago  Section 
in  the  Auditorium  of  the  Western  So- 
ciety of  Engineers,  June  27,  Mr.  Arthur 
J.  Sweet  presented  a  paper  entitled 
"Notes  on  Postal  Car  Illumination." 

The  following  officers  have  been 
elected  for  the  year  beginning  October 
1,  1913:  chairman,  Dr.  M.  G.  Lloyd; 
secretary,  J.  B.  Jackson;  managers,  J. 
W.  Pfeifer,  Dr.  Nelson  M.  Black,  C.  C. 
Schiller,  M.  J.  Sturm  and  H.  B. 
Wheeler.  Mr.  Jackson  was  re-elected 
secretary. 

NEW    ENGLAND    SECTION. 

The  following  officers  have  been 
elected  for  the  ensuing  year :  chairman, 
C.  A.  B.  Halvorson,  Jr.;  secretary,  H. 
Harold  Higbie ;  managers,  R.  B.  Hus- 
sey,  H.  C.  Jones,  J.  M.  Riley,  R.  C. 
Ware  and  W.  E.  Wickenden.  Since  the 
election,  Prof.  Higbie  has  removed  from 
the  territory  of  the  New  England  and 
has  tendered  his  resignation  as  secretary. 
A  secretary  will  be  appointed  by  the 
new  section  board. 

NEW    YORK    SECTION. 

The  following  officers  have  been 
elected  for  the  ensuing  year :  chairman, 
W.  Cullen  Morris;  secretary,  Clarence 
L.  Law;  managers,  H.  B.  Rogers,  C.  R. 
Clifford,  H.  V.  Allen,  W.  H.  Spencer 
and  Oscar  Fogg.  Mr.  Law  was  re- 
elected secretary. 


TRANSACTIONS   I.    E.    S. — PART    I 


PHILADELPHIA    SECTION. 

The  Philadelphia  Section  held  a  din- 
ner at  the  Engineers'  Club,  1317  Spruce 
Street,  Friday  evening,  June  20.  The 
dinner  was  held  in  place  of  the  regular 
June  meeting  and  was  attended  by 
twenty-five  members.  Prof.  James 
Barnes  of  Bryn  Mawr  College  gave  a 
short  talk  on  "Recent  Ideas  in  Regard 
to  the  Spectrum."  Dr.  Wendell  Reber 
gave  a  short  address  on  "The  Problem 
of  the  Relation  of  the  Human  Eye  to 
Illumination,   Natural  and  Artificial." 

The  following  officers  have  been 
elected  for  the  year  beginning  October 
1  :  chairman,  Prof.  George  A.  Hoadley ; 
secretary,  L.  B.  Eichengreen ;  managers, 
F.  C.  Dickey,  H.  H.  Ganzer,  H.  A. 
Hornor,  H.  Calvert  and  Samuel  Snyder. 
Mr.  Eichengreen  was  re-elected  secre- 
tary. 

PITTSBURGH    SECTION. 

At  a  meeting  of  the  Pittsburgh  Sec- 
tion in  the  auditorium  of  the  Engineers' 
Society  of  Western  Pennsylvania, 
Oliver  Building,  June  20,  Messrs.  Evan 
J.  Edwards  and  Ward  Harrison  of  the 
National  Electric  Lamp  Association 
presented  a  paper  entitled  "Some  Engi- 
neering Features  of  Office  Lighting." 
The  paper  will  be  published  in  the  next 
issue  of  the  Transactions,  the  October 
number. 

Announcement  was  made  of  the  elec- 
tion of  the  following  officers  for  the 
ensuing  year:  chairman,  C.  J.  Mundo ; 
secretary,  Alan  Bright ;  managers,  H.  S. 
Hower,  H.  H.  Magdsick,  E.  R.  Roberts, 
C.  E.  Stephens  and  S.  B.  Stewart. 


October  1  :  president,  Charles  O.  Bond ; 
vice-president  to  represent  the  New 
York  Section,  George  H.  Stickney; 
vice-president  to  represent  the  Pitts- 
burgh Section,  Ward  Harrison ;  general 
secretary,  Joseph  D.  Israel ;  treasurer, 
L.  B.  Marks ;  directors,  F.  J.  Rutledge, 
C.  A.  Littlefield,  F.  A.  Vaughn.  Di- 
rectors are  each  elected  for  three 
years,  vice-presidents,  two  years,  and 
the  other  officers  one  year.  Messrs. 
Israel  and  Marks  were  re-elected. 

The  results  of  the  election  for  officers 
of  the  several  sections  may  be  found 
under  the  Section  Notes  in  this  issue. 


New  Officers. 

At  the  recent  annual  election  of  the 
Society  the  following  officers  were 
elected     for    various    terms    beginning 


Charles  O.  Bond,  president-elect  of 
the  Illuminating  Engineering  Society, 
was  born  November  15,  1870,  near  the 
town  of  Lehigh,  Webster  County,  Iowa. 
He  was  graduated  from  the  United 
States  Naval  Academy,  Annapolis,  Md. 
in  1890.  After  graduation  he  served 
one  year  at  sea  on  board  the  U.  S.  S. 
Enterprise  and  U.  S.  S.  Philadelphia, 
resigning  from  the  navy  in  1891.  He 
later  taught  school  in  the  states  of 
Iowa  and  New  York  for  five  years,  and 
in  1897  became  connected  with  the 
United  Gas  Improvement  Company  in 
Philadelphia.  In  1898,  during  the 
Spanish-American  War,  he  served  five 
months  as  an  ensign  in  the  navy  on 
board  the  U.  S.  S.  Lancaster  and  the 
U.  S.  S.  Newport.  At  the  close  of  the 
war,  he  resumed  his  connection  with  the 
United  Gas  Improvement  Company,  tak- 
ing charge  of  the  photometric  work  of 
the  company.  While  continuing  in  this 
position,  he  also  held  command  of  one 
division  of  "the  Naval  Force  of  Pennsyl- 
vania for  three  years.  Since  1909,  Mr. 
Bond  has  been  manager  of  the  photo- 
metrical  laboratory  of  the  United  Gas 
Improvement  Company,  which  was  es- 
tablished in  that  year. 


4 


TRANSACTIONS    I.    E.    S. — PART    I 


Joseph  D.  Israel,  who  has  twice  been 
elected  general  secretary  of  the  society, 
has  been  connected  with  the  lighting 
industry  during  the  past  twenty-six 
years.  He  is  at  present  district  manager 
of  the  Philadelphia  Electric  Company. 

Mr.  Israel  was  born  in  Philadelphia, 
February  28,  1868.  In  1886  he  was 
graduated  from  the  Scientific  School  of 
the  University  of  Pennsylvania  with  the 
degree  of  Bachelor  of  Science.  After 
a  post-graduate  year  in  mechanical  and 
electrical  courses,  the  same  university 
conferred  upon  him  the  degree  of  Me- 
chanical Engineer  in  1887.  Immediately 
after  graduation  he  became  connected 
with  the  Edison  Electric  Light  Com- 
pany of  Philadelphia,  devoting  his  time 
to  underground  street  work.  Shortly 
afterward  he  became  superintendent  of 
the  street  work  of  the  company.  He 
next  became  assistant  to  the  manager 
of  the  company,  and  later  was  made 
secretary  and  manager.  When  the  com- 
pany was  merged  with  the  Philadelphia 
Electric  Company  he  became  district 
manager  of  the  latter  company. 

Mr.  Israel  has  been  active  in  the  local 
and  national  work  of  the  Illuminating- 
Engineering  Society,  the  National  Elec- 
tric Light  Association  and  the  American 
Institute  of  Electrical  Engineers.  He 
is  also  a  member  of  the  Franklin  Insti- 
tute of  the  State  of  Pennsylvania  and  a 
director  of  the  Commercial  Section  of 
the  National  Electric  Light  Association. 
He  has  contributed  papers  and  reports 
to  local  and  national  meetings  of  the 
above-named  societies,  and  the  Associa- 
tion of  Edison  Illuminating  Companies. 


A  Survey  of  Present  Day  Lighting. 

The  Illuminating  Engineering  So- 
ciety, through  its  president,  Mr.  Preston 
S.  Millar,  is  undertaking  the  preparation 


of  an  exhaustive  survey  of  present  day 
lighting  conditions,  the  object  being  to 
record  as  nearly  as  possible  the  lighting 
practise  in  a  number  of  different  fields. 
It  is  expected  that  the  survey  will  af- 
ford concrete  information  on  the  pres- 
ent situation  and  provide  a  basis  of 
comparison  for  future  estimates  of 
progress.  It  will  probably  be  of  most 
value  at  the  present  time  in  enabling 
individuals  and  companies  to  compare 
their  practises  with  general  practise 
and  conditions ;  it  should  disclose  what 
is  judged  the  most  advanced  practise  in 
each  field. 

For  the  purpose  of  this  survey  the 
United  States  has  been  divided  into 
eighty-nine  different  sections,  each  sec- 
tion containing  approximately  one  mil- 
lion inhabitants.  Into  each  district  vari- 
ous lists  of  questions  regarding  illumi- 
nating practise  are  being  sent  to  repre- 
sentative companies  and  individuals  in 
the  following  professions  and  indus- 
tries : 

Central  stations. 
Gas  companies. 
Municipal  engineers. 
Manufacturers  of  incandescent  lamps. 
Manufacturers  of  mantle  burner  lamps. 
Manufacturers  of  arc  lamps. 
Manufacturers  of  acetylene  supplies,  tips,  etc. 
Manufacturers  of  oil  lamps. 
Manufacturers  of  small  isolated  lighting  plant 

equipments,  gasolene,  acetylene,  etc. 
Manufacturers  of  lighting  glassware. 
Fixture  manufacturers. 
Arc  lamp  post  manufacturers. 
Ophthalmologists. 

School  associations  and  commissions. 
Railroads. 

Street  railroad  companies. 
Street  lighting  lamp  companies. 

It  is  hoped,  therefore,  that  with  a 
reasonable  amount  of  co-operation  on 
the  part  of  those  to  whom  the 
questions    are    sent   that    estimates    and 


""      C<.' 


S 


CHARLES  O.  BOND,  President-elect. 


JOSEPH  D.  ISRAEL,  General  Secretary. 


\ 


TRANSACTIONS    I.    E.    S. — PART    I 


data  will  be  received  from  some  two 
to  three  thousand  persons,  each  of 
whom  is  probably  best  fitted  to  pro- 
vide the  information  for  which  he  is 
asked.  In  interpreting  and  summariz- 
ing the  information,  which  it  is  hoped 
will  be  made  available,  an  effort  will  be 
to  emphasize  that  which  is  constructive 
and  to  avoid  everything  invidious.  The 
final  survey  will  be  made  up  with  the 


co-operation  and  advise  of  a  number  of 
men  in  the  lighting  industry,  and  no 
information  will  be  used  which  will  be 
likely  to  prove  in  any  way  derogatory 
to  the  interests  of  those  contributing. 
The  information  derived  will  be  used  as 
a  basis  of  a  paper  to  be  presented  at  the 
convention  of  the  Illuminating  Engi- 
neering Society  in  Pittsburgh  during 
the  week  beginning  September  22,  1913. 


6  TRANSACTIONS   I.    E.    S. — PART   I 

ANNOUNCEMENT    BY    COMMITTEE    ON   NOMENCLA- 
TURE AND  STANDARDS. 


The  committee  has  tentatively  adopted  the  following  addi- 
tional definitions : 

Apparent  candle-power,  at  a  distance  d,  is  the  candle-power  of 
the  simple  luminous  source  which  at  the  distance  d  from  the 
point  of  observation  would  give  an  illumination  equal  to  the  ob- 
served illumination. 

The  term  should  be  used  only  for  cases  in  which  the  law  of 
inverse  squares  does  not  apply.  The  term  is  meaningless  unless 
d  is  given. 

Power  consumed  by  a  source,  P,  expressed  in  ergs  per  second  is 
the  total  power  input  in  a  radiating  body. 

Power    radiated  by   a    source,    Pr   =    1  PrA   dK,     the    radiant 

power  emitted  by  a  source  in  the  form  of  radiation  between  wave- 
lengths zero  and  infinity,  expressed  in  ergs  per  second. 

pr 
Radiation  Efficiency,  p  —  — ,  a   numeric,    is  the  ratio  of  the 

power  radiated  to  the  total  power  consumed  by  a  source. 

Specific  Consumption,  the  ratio  of  the  power  consumed  by  a 
source  to  the  total  luminous  flux  in  lumens. 

Specific  luminous  output,  the  reciprocal  of  specific  consump- 
tion, F/P. 

The  committee  further  would  present  to  the  membership  the 
following  definitions  proposed  by  Dr.  H.  E.  Ives  and  not  yet 
acted  upon  by  the  committee.  The  committee  desires  the  sug- 
gestions and  criticisms  of  the  membership  on  all  of  the  definitions 
here  given. 

po.S/u 

Available  useful  power  for  lighting  purposes,  Re  =  I  RA  d\  = 

Jo.4fi 

the  radiation  lying  in  the  visible  spectrum.  (The  power  which 
gives  one  mean  spherical  candle  of  this  radiation  is  sometimes 
called  the  mechanical  equivalent  of  a  light). 

Visible  fraction  of  the  radiation  from  a  source  =  ~~  =  fraction 
of  the  total  radiation   useful   for   lighting   purposes.     (A  pure 


TRANSACTIONS   I.    E.    S. — PART    I 


numeric).   (This  fraction  is  sometimes  called  the  radia?it  luminous 
efficiency,  the  ratio    -e ,   the   total  luminous   efficiency  of  a    light 

source). 

R  j.      * 

Radiant  specific  consumption   =  -=r  =  power    radiated     per 

lumen  =  p-^r- 

F 
Radiant  specific  luminous  output  —=-  =    lumens    radiated    per 

F 
watt  radiated  =  Km=  -=■  (.*.  identical  with  stimulus  coefficient) . 
p¥ 

Note:— Watts  per  candle,  candles  per  watt,   watts  per  lumen,  lumens 
per  watt,  are  ratios  sometimes  called  "  efficiency." 

Radiant  Luminous  Efficie?icy  =   fxR    =   ratio  of  the  radiant 
specific  luminous  output  of  a  source  to  the  maximum  possible 
specific  luminous  output.      (A  pure  numeric) 
JF 
R  Km 


Pr 


(i), 


i      rc 

'  max  J 


KA  Ra  a\. 


Total  luminous  efficiency  =  /»  =  ratio  of  the  specific  luminous 
output  of  a  source  to  the  maxium  possible  specific  luminous 
output.     (A  pure  numeric) 

P  R         _  K^p 

K„ 


(f  L  (1) 


max 
max 


Since  maxium  value  of  p  is  unity. 

Note  : — If  Kmax  be  taken  as  unity,  i.  e.,  if  the  unit  of  flux  be  taken  as 
that  given  by  one  power  unit  of  radiation  of  maxium  luminous  efficiency, 
then  considering  the  radiation  from  a  source 

F         _ 

MR  =  -g-  =  Kg, 

or  luminous  efficiency  =  specific  luminous  output  =  stimulus  coefficient 
(numerically). 

For  the  total  radiation 

F  _ 

or  luminous  efficiency  is  numerically  the  same  as  specific  consumption. 

C.   H.  Sharp, 
Secretary  Committee  on  Nomenclature  and  Standards. 


3  TRANSACTIONS    I.    E.    S. — PART    I 

Constitution  and  By-laws 

OF  THE 

Illuminating  Engineering  Society 

(Adopted  by  a  vote  of  the  Membership,  January  14,  1907.) 

Constitution  amended:  Jan.,   1909;  Jan.,  1910;  Jan.,  1911;  Jan.,  1912;  Dec, 

1912.     By-laws  amended  by  Council:    Jan.    28,    1907;    Feb.   10,    1910; 

Mar.  10,  1911  ;  Dec.  8,  1911;  Jan.  12,  1912;  Jan.  10,  1913. 

ARTICLE  I. 

NAME    AND    OBJECTS. 

Section  1 :  The  name  of  this  association  shall  be  the  Illumi- 
nating Engineering  Society. 

Section  2:  Its  objects  shall  be  the  advancement  of  the  theory 
and  practise  of  illuminating  engineering  and  the  dissemination 
of  knowledge  relating  thereto.  Among  the  means  to  this  end 
shall  be  meetings  for  the  presentation  and  discussion  of  appro- 
priate papers ;  the  publication  as  may  seem  expedient  of  such 
papers,  of  discussions  and  communications ;  and  through  com- 
mittees, the  study  of  subjects  relating  to  the  science  and  art  of 
illumination,  and  the  publication  of  reports  thereon. 

(a)  Sec.  2:  The  appointment  of  committees  to  report  upon 
scientific  and  engineering  subjects  shall  be  authorized  only  by 
a  majority  vote  of  the  Council,  which  shall  be  taken  by  letter- 
ballot.  When  such  a  committee  is  thus  authorized,  the  President 
shall  appoint  the  members  thereof,  subject  to  approval  by  vote  of 
a  quorum  of  the  Council. 

ARTICLE  II. 

MEMBERSHIP. 

Section  1 :  The  members  of  this  Society  shall  be  designated 
as  Members,  Sustaining  Members  and  Honorary  Members. 

Section  2:  A  Member  may  be  anyone  interested  in  the  objects 
of  the  Society.  At  the  time  of  his  election  he  shall  not  be  less 
than  twenty-one  years  of  age. 

By-laws  s  re  printed  in  small  type. 


TRANSACTIONS    I.    E.    S.—  PART    I  9 

Section  3:  A  Sustaining  Member  may  be  a  company,  firm, 
association  or  individual  interested  in  the  objects  of  the  Society 
and  desirous  of  contributing  to  its  support.  A  Sustaining  Mem- 
ber, when  other  than  an  individual,  may  be  officially  represented 
by  an  individual.  The  privileges  of  Sustaining  Members  shall 
be  the  same  as  those  of  members,  except  the  right  to  vote  and 
to  hold  office.  All  provisions  of  this  Constitution  governing  the 
admission,  duties  and  obligations  of  members  shall,  unless  other- 
wise provided,  apply  to  Sustaining  Members. 

Section  4:  Honorary  Members  may  be  chosen  from  among 
those  who  are  of  acknowledged  eminence  in  some  branch  of  art 
or  science  related  to  illuminating  engineering.  Honorary  Mem- 
bers shall  be  entitled  to  all  the  privileges  of  the  Society  except 
the  right  to  vote  and  to  hold  office  therein. 

ARTICLE  III. 

ADMISSION    AND    EXPULSION    OF    MEMBERS. 

Section  1 :  Honorary  Members  shall  be  proposed  in  writing 
by  at  least  fifteen  members,  and  shall  be  elected  only  by  the 
unanimous  vote  of  the  Council.  Voting  shall  be  by  letter-ballot. 
A  person  elected  an  Honorary  Member  shall  be  promptly  noti- 
fied by  letter,  and  the  election  shall  be  cancelled  if  an  acceptance 
is  not  received  within  six  months  after  the  mailing  of  such 
notice. 

Section  2:  An  application  for  admission  to  the  Society  shall 
be  made  in  a  form  prescribed  by  the  Council  and  shall  bear  the 
endorsement  of  at  least  two  Members  of  the  Society;  or  shall 
refer  to  at  least  two  Members  of  the  Society;  or  if  an  applicant 
certifies  that  he  is  not  personally  known  to  two  members,  refer- 
ences may  be  accepted  to  members  of  professional  societies  of 
good  standing,  or  to  other  persons  whose  good  standing  may  be 
readily  verified. 

(a)  Sec.  2:  An  application  for  membership  in  the  Society 
shall  be  made  upon  a  printed  form  prepared  by  the  General  Secre- 
tary and  approved  by  the  Council,  which  shall  call  for  such  in- 
formation as  may  be  required  by  the  Board  of  Examiners  and 
the  Council  to  pass  properly  upon  the  eligibility  of  a  candidate. 

By-laws  are  printed  in  small  type. 


10  TRANSACTIONS    I.    E.    S. — PART   I 

(b)  Sec.  2:  In  the  absence  of  replies  from  referees  to  in- 
quiries for  information,  or  if  replies  are  not  sufficiently  explicit, 
the  Board  of  Examiners  having  cognizance  of  the  application 
shall  cause  the  applicant  to  be  notified  and  shall  hold  his  appli- 
cation in  abeyance. 

Section  3:  All  applications  for  admission  to  membership  shall 
be  passed  upon  by  a  Board  of  Examiners  of  the  section  of  the 
Society  representing  the  locality  in  which  the  applicant  resides. 
All  applications  shall  be  reported  to  the  Council  for  final  action. 
An  applicant  not  residing  within  the  territory  of  a  section  shall 
submit  his  application  direct  to  the  Council. 

(c)  Sec.  3 :  When  applications  for  admission  are  received 
from  persons  residing  within  the  territory  of  a  section  the 
General  Secretary  shall  notify  the  Secretary  of  that  section  to 
make  prompt  report  upon  the  application. 

(d)  Sec.  3:  The  privileges  attaching  to  membership  in  the 
Society  shall  not  be  accorded  to  newly-elected  members  until 
they  have  paid  their  entrance  fee  and  current  dues.  This  by-law 
shall  be  printed  upon  the  notification  of  election. 

(e)  Sec.  3:  Upon  receipt  of  an  application  for  membership, 
which  shall  be  made  on  the  official  form  the  General  Secretary, 
or  the  Secretary  of  a  section,  shall  see  if  it  has  been  properly 
filled  out.  If  not,  he  shall  return  the  form  and  notify  the  appli- 
cant of  the  deficiency.  When  an  application  is  in  proper  form 
it  shall  be  forwarded  to  the  chairman  of  the  Board  of  Examiners. 
The  Secretary  of  a  section  shall  conduct  for  the  Board  of  Ex- 
aminers such  correspondence  with  applicants  and  their  referees 
as  the  Board  may  direct. 

(f)  Sec.  3:  Objection  to  the  admission  of  a  candidate  must  be 
accompanied  by  specific  reasons  for  such  objection. 

Section  4:  A  Member  may  resign  from  the  Society  by  a 
written  communication  to  the  Secretary,  which  resignation  shall 
be  accepted  by  the  Council  if  all  his  dues  and  other  indebtedness 
have  been  paid,  and  the  Society  badge  has  been  returned. 

Section  5:  Upon  the  written  request  of  ten  or  more  members 
that,  for  cause  definitely  stated  in  detail,  a  Member  of  the  So- 
ciety be  expelled,  the  Board  of  Managers  of  /the  section  of  the 
locality  wherein  the  accused  resides  shall  consider  the  matter, 
and  if  there  appears  to  be  sufficient  cause  shall  advise  the  ac- 
cused of  the  charges  against  him.     The  accused  may  then  present 

By-laws  are  printed  in  small  type. 


TRANSACTIONS    I.    E.    S. — PART    I  11 

a  written  defense  and  appear  in  person  before  a  meeting  of  the 
board.  The  finding  of  the  board  shall  then  be  submitted  to  the 
Council  of  the  Society  which,  within  two  months,  shall  finally 
consider  the  case,  and  if  a  satisfactory  defense  has  not  been 
made,  the  accused  Member  shall  be  expelled  upon  a  two-thirds 
vote  of  the  Council.  In  the  case  of  one  not  a  member  of  a 
section,  charges  shall  be  preferred  directly  to  the  Council. 

ARTICLE  IV. 

DUES. 

Section  1:     An   entrance    fee,    payable   on    admission   to   the 
Society,  may  be  fixed  by  the  Council. 

(a)  Sec.  i :  The  entrance  fee  for  members  shall  be  $2.50  pay- 
able on  admission  to  the  Society.  There  shall  be  no  entrance  fee 
for  Sustaining  Members 

Section  2:     The  annual  dues  for  members  shall  be  $5,  which 
shall  include  subscription  to  the  Transactions  of  the  Society. 

(b)  Sec.  2:  The  annual  dues  are  payable  in  advance.  Bills 
for  dues  shall  be  sent  out  by  the  General  Secretary  not  later 
than  October  10. 

(c)  Sec.  2:  If  the  entrance  fee  and  dues  are  not  paid  within 
one  month  after  a  member  has  been  notified  of  his  election,  he 
shall  be  finally  informed  of  the  delinquency;  and  if  such  dues 
are  not  paid  within  two  months  from  the  time  of  notification  of 
election,  the  Council  shall  cancel  the  election,  of  which  cancella- 
tion the  delinquent  and  his  referees  shall  be  informed.  This  by- 
law shall  be  printed  on  the  final  notice  above  provided  for.  (See 
also  by-law  Sec.  3,  Art  III.) 

(d)  Sec.  2:  Any  member  in  arrears  four  months  for  dues, 
shall  be  informed  by  the  General  Secretary  that  he  is  delinquent 
and  can  have  no  vote  or  voice  in  the  affairs  of  the  Society  or 
receive  its  Transactions  or  other  publications  until  the  dues  are 
paid.  At  the  expiration  of  two  months  thereafter,  if  still  in  ar- 
rears, he  shall  be  notified  that  his  name  will  be  presented  to  the 
Council  as  delinquent,  if  the  dues  are  not  paid  within  one  month. 
If  the  member  continues  delinquent,  the  Council  shall  drop  him 
from  membership  at  the  regular  meeting  held  in  June. 

(e)  Sec.  2:     From  the  annual  dues  paid  by  each   Member  $3 
shall  be  deducted  and  applied  as  a  subscription  to  the  Transac- 
tions for  the  year  covered  by  such  payment.     The  price  of  sub- 
By-laws  are  printed  in  small  type. 


12  TRANSACTIONS    I.    E.    S.— PART    I 

scription  of  the  Transactions  to  non-members  of  the  Society 
shall  be  $5  per  year.  Single  copies  may  be  sold  at  55  cents  each ; 
provided  that  volumes  reserved  shall  not  be  broken  to  furnish 
single  copies. 

(f)  Sec.  2:  The  official  badge  of  the  Society  shall  be  issued 
by  the  General  Secretary  upon  application  to  any  Member  in 
good  standing,  upon  payment  of  $3 ;  provided  that  Honorary  Mem- 
bers shall  receive  the  badge  without  payment.  Each  badge  shall 
be  numbered  and  registered  in  the  name  of  the  member  receiving 
it.  Members  purchasing  badges  shall  be  informed  by  the  Gen- 
eral Secretary  that  they  are  issued  with  the  express  condition 
that  if  the  member  resigns  or  is  dropped  from  the  Roll  of  the  So- 
ciety, he  shall  return  his  badge,  receiving  therefor  the  sum  of  $2. 

(g)  Sec.  2 :  A  certificate  of  membership  in  the  Society  shall 
be  issued  by  the  General  Secretary  upon  application  to  any  mem- 
ber in  good  standing,  upon  payment  of  $1.00. 

Section  3:  The  annual  dues  for  Sustaining  Members  shall  be 
not  more  than  $250.00. 

(h)  Sec.  3:     Dues   (not  exceeding  $250  annually)   for  Sustain- 
ing   Members    may    be    elective    with    the    Sustaining    Member. 
Transactions  shall  be  sent  free  of  charge  to  Sustaining  Mem- 
bers whose  dues  are  $10.00  or  more  per  annum. 
Section  4:     Honorary  Members  shall  be  exempt  from  all  pay- 
ments. 

Section  5:  A  Member  elected  after  six  months  of  the  fiscal 
year  have  expired  shall  pay  one-half  of  the  amount  of  dues  for 
that  year;  provided,  that  if  he  requests  and  receives  a  set  of 
Transactions  covering  the  entire  year,  then  the  full  annual  dues 
shall  be  paid. 

Section  6:     A  Member  who  has  been  dropped  as  delinquent 
may  be  reinstated  by  the  Council  and  retain  his  original  date  of 
election  upon  payment  of  all  back  dues,  being  then  entitled  to  a' 
complete  file  of  the  publications  of  the  Society,  if  in  stock,  cor- 
responding to  the  period  of  delinquency. 

ARTICLE  V. 

OFFICERS. 

Section  1:  The  officers  of  the  Society  shall  be  a  President. 
Vice-presidents  equal  in  number  to  the  number  of  organized 
Sections,  nine  Directors,  a  Secretary  and  a  Treasurer. 


By-laws  are  printed  in  small  type. 


TRANSACTIONS    I.    E.    S.  —  PART    I  13 

Section  2:  The  President,  the  Secretary  and  the  Treasurer 
shall  hold  office  for  one  year ;  the  Vice-presidents  shall  hold  office 
for  two  years  and  the  Directors  for  three  years.  Terms  of  of- 
fice shall  commence  the  first  day  of  October.  A  retiring  Presi- 
dent, Vice-president  or  Director  shall  not  be  eligible  for  imme- 
diate re-election  to  the  same  office,  and  a  retiring  Vice-president 
shall  not  be  eligible  for  immediate  election  as  a  Director.  At 
each  annual  meeting  officers  shall  be  elected  to  succeed  those 
retiring  by  expiration  of  term. 

Section  3:  A  vacancy  in  the  office  of  President  shall  be  filled 
by  the  senior  Vice-president ;  a  vacancy  in  the  office  of  Vice- 
president  shall  be  filled  by  the  senior  Director ;  a  vacancy  in  the 
office  of  Director  shall  be  filled  by  the  Council,  preferably  by 
selection  from  members,  if  any,  who  at  the  previous  annual 
election  received  votes  for  the  office  of  Director.  A  vacancy  in 
the  office  of  Secretary  or  Treasurer  shall  be  filled  by  the  Council. 
Such  succession  to  office  or  appointment  by  the  Council  shall  not 
render  an  officer  ineligible  for  immediate  election  to  the  same 
office.  Seniority  between  officers  of  the  same  rank  and  date  of 
election  shall  be  determined  by  the  date  of  their  election  as  mem- 
bers. 

Section  4:  No  officer  shall  receive,  directly  or  indirectly,  any 
salary,  compensation  or  emolument  from  the  Society,  either  as 
such  officer  or  in  any  other  capacity,  unless  authorized  by  a  vote 
of  the  majority  of  the  entire  Council.  No  officer  shall  be  inter- 
ested, directly  or  indirectly,  in  any  contract  relating  to  the  opera- 
tions conducted  by  the  Society,  nor  in  any  contract  for  furnishing 
supplies  thereto,  unless  by  the  unanimous  vote  of  the  Council. 

ARTICLE  VI. 

ELECTION   OF  OFFICERS. 

Section  1:  Each  year  not  later  than  April  I,  a  Board  of 
Nomination,  consisting  of  the  two  junior  Past-presidents  and  of 
the  Past-vice-presidents  whose  terms  of  office  expired  in  the  two 
preceding  Septembers,  shall  proceed  to  prepare  a  nomination 
ticket  containing  the  names  of  those  whom  they  deem  best  suited 
for  the  offices  to  be  filled  at  the  ensuing  annual  election.     Nomi- 

2 


14  TRANSACTIONS    I.    E.    S. — PART   I 

nees  for  the  office  of  Vice-president  shall  be  so  selected  that,  if 
such  nominees  are  elected,  each  locality  where  there  is  a  Section 
of  the  Society  may  be  represented  on  the  Council  by  a  Vice- 
president. 

(a)  Sec.  i :  Each  year,  not  later  than  March  15,  the  General 
Secretary  shall  send  to  the  senior  Past-Officer  among  those  desig- 
nated in  Article  VI,  Section  i,  a  copy  of  that  section  and  the 
names  and  addresses  of  the  other  Past-Officers  therein  designated. 
The  senior  officer  shall  then  forthwith  proceed  to  organize  the 
Board  of  Nomination  and  shall  submit  its  report  to  the  General 
Secretary  not  later  than  April  25. 

Section  2:  The  ticket  thus  prepared  shall  be  printed  and  for- 
warded to  members  not  later  than  May  5  together  with  an  un- 
marked inner  envelope,  and  an  outer  official  voting  envelope 
bearing  the  name  and  address  of  the  Society  and  the  words, 
"Official  Voting  Envelope — Enclosing  a  Ballot  Only."  The 
member  voting  shall  enclose  his  ballot  in  the  unmarked  envelope, 
which  shall,  in  turn,  be  enclosed  in  the  outer  envelope,  which 
latter  shall  be  endorsed  with  the  name  of  the  sender.  Ballots  to 
be  counted  must  reach  the  General  Secretary  not  later  than 
May  26. 

(b)  Sec.  2 :  The  General  Secretary  shall  have  printed  and 
enclose  with  the  official  nomination  ticket,  Section  2  of  Article 
V,  and  Sections  1,  2  and  3  of  Article  VI. 

(c)  Sec.  2:  The  roll  of  members  shall  designate  those  who 
are  charter  members  of  the  Society.  The  names  of  present  and 
past  general  officers  shall  be  followed  by  the  name  of  office  held, 
printed  in  italic  type. 

Section  3:  A  member  may  vote  the  official  ticket  above  pro- 
vided for;  or  he  may  erase  any  names  thereon  and  substitute 
others;  or  he  may  substitute  a  written  ballot  containing  names 
of  his  own  selection. 

Section  4:  The  President  at  a  Council  meeting  in  May  shall 
appoint,  subject  to  the  approval  of  the  Council,  five  members,  not 
members  of  the  Council,  to  constitute  a  Committee  of  Tellers. 
This  committee  shall  meet  between  May  26  and  May  30,  and 
shall  receive  unopened  all  ballots  from  the  General  Secretary  and 
shall  forthwith  proceed  in  secret  to  count  the  vote.    It  shall  then 

By-laws  are  printed  in  small  type. 


TRANSACTIONS    I.    E.    S. — PART   I  15 

prepare  in  duplicate  and  sign  a  report  of  the  results  of  the  vote, 
one  copy  of  which  shall  be  delivered  to  the  General  Secretary  and 
the  other  handed  by  the  chairman  of  the  committee,  at  the  en- 
suing annual  meeting,  to  the  presiding  officer,  who  shall  at  the 
opening   session   of   the   meeting   announce   the   names   of   the 

officers  elected. 

(d)  Sec.  I :  The  General  Secretary  upon  receipt  of  the  report 
from  the  Committee  of  Tellers,  shall  at  once  notify  the  president- 
elect of  the  result  of  the  election. 

ARTICLE  VII. 

MANAGEMENT. 

Section  1:  The  affairs  of  the  Society  shall  be  managed  by  a 
Council  under  this  Constitution  and  under  the  By-Laws  adopted 
for  the  execution  thereof.  The  Council  shall  direct  the  business 
of  the  Society  either  itself  or  through  its  officers  and  com- 
mittees. 

Section  2:  The  Council  shall  consist  of  the  officers  of  the 
Society  and  of  the  two  junior  Past-presidents. 

(a)  Sec.  2:  Regular  meetings  of  the  Council  shall  be  held 
once  each  month,  except  during  July,  August  and  September. 
Special  meetings  of  the  Council  or  of  the  Executive  Committee 
may  be  called  by  the  President.  Notice  of  such  special  meetings 
shall  be  forwarded  to  the  members  of  the  Council  or  of  the 
Executive  Committee  at  least  three  days  in  advance  of  the 
meeting.  The  notice  shall  contain  a  synopsis  of  the  business 
to  be  brought  before  the  special  meeting,  and  no  business  other 
than  that  so  specified  shall  be  transacted  at  such  meeting. 

(b)  Sec.  2:  The  General  Secretary  shall,  after  each  meeting 
of  the  Council,  forward  to  each  member  thereof  a  transcript  of 
the  minutes  of  the  meeting. 

Section  3 :  The  Council  may  delegate  any  or  all  of  its  powers 
to  an  Executive  Committee  of  five  members,  consisting  of  the 
President,  the  Secretary  and  the  Treasurer,  ex-officio,  and  two 
other  members  of  the  Council,  which  committee  shall  conduct  the 
affairs  of  the  Council  between  its  meetings. 

(c)  Sec.  3 :  Should  the  Executive  Committee  have  taken  any 
action  between  meetings  of  the  Council,  it  shall  report  such 
action  at  the  first  meeting  of  the  Council  'following;  if  approved, 
the  action  of  the  Executive  Committee  shall  be  as  if  the  action 
of  the  Council. 


By-laws  are  printed  in  small  type. 


16  TRANSACTIONS    I.    E.    S. — PART    I 

Section  4:  The  President  shall  have  general  supervision  of 
the  affairs  of  the  Society  under  the  direction  of  the  Council.  He 
shall  preside  at  the  meetings  of  the  Council  at  which  he  may  be 
present  and  shall  be  ex-officio  member  of  all  committees.  He 
shall  deliver  an  address  at  the  annual  convention  of  the  Society. 

Section  5:  Vice-presidents  or  Directors,  in  order  of  seniority, 
shall  preside  at  meetings  of  the  Council  in  the  absence  of  the 
President. 

Section  6:  The  Treasurer  shall  be  the  custodian  of  all 
moneys.  He  shall  make  an  annual  report,  which  shall  be  au- 
dited, and  such  other  reports  as  may  be  prescribed.  The  Treas- 
urer and  the  Secretary,  with  the  advice  and  consent  of  the  Com- 
mittee on  Finance,  shall  invest  such  funds  as  may  be  ordered  by 
the  Council.  They  shall  pay  all  bills  when  audited  by  the  Com- 
mittee on  Finance  and  approved  by  the  Council. 

Section  7:  The  General  Secretary  shall  be,  under  the  direc- 
tion of  the  President  and  the  Council,  the  executive  officer  of  the 
Society.  He  shall  prepare  the  business  for  the  Council  and  re- 
cord the  proceedings  thereof.  He  shall  collect  all  moneys  due 
to  the  Society,  and  deposit  the  same  subject  to  the  order  of  the 
Treasurer.  He  shall  personally  certify  the  accuracy  of  bills  or 
vouchers  upon  which  money  is  to  be  paid  and  shall  draw  and 
countersign  all  checks,  which  shall  be  signed  by  the  Treasurer 
when  such  drafts  are  known  by  him  to  be  proper,  duly  authorized 
by  the  Committee  on  Finance  and  in  accordance  with  the  neces- 
sary vouchers  transmitted  by  the  General  Secretary  with  the 
draft.  He  shall  have  charge  of  the  books  and  accounts  of  the 
Society  and  shall  furnish  monthly  to  the  Council  a  state- 
ment of  receipts  and  expenditures  and  monthly  balances.  He 
shall  present  annually  a  report  to  the  Council  for  publication  in 
the  Transactions,  and  from  time  to  time  shall  furnish  such 
statements  as  may  be  required.  He  shall  conduct  the  corre- 
spondence of  the  Society  and  keep  full  records  and  perform  such 
other  duties  as  may  be  assigned  to  him.  The  Council  may  ap- 
point assistants  to  the  General  Secretary ;  one  of  these  may  have 
the  title  of  Assistant  Secretary,  and  shall  be  under  the  immediate 
direction  of  the  General  Secretary  and  aid  him  in  all  matters. 


>? 


TRANSACTIONS    I.    E.    S.  —  PART    I 

In  the  event  of  prolonged  absence  or  disability  of  the  General 
Secretary  or  Treasurer  the  Council  shall  authorize  one  of  its 
members  to  sign  or  countersign  checks. 

(d )  Sec.  7 :  The  accounts  of  the  General  Secretary  and  the 
Treasurer  shall  be  audited  annually  just  prior  to  the  annual 
meeting. 

Section  8:  The  President  shall,  at  the  first  meeting  of  the 
Council  after  he  assumes  office,  appoint,  subject  to  the  approval 
of  the  Council,  the  following  standing  committees :  a  Commit- 
tee on  Finance,  of  three  members;  a  Committee  on  Papers,  of  at 
least  five  members ;  a  Committee  on  Editing  and  Publication,  of 
three  members.  He  may  also  appoint  temporary  committees 
from  time  to  time.  Two  of  the  three  members  of  the  Finance 
Committee  shall  be  members  of  the  Council,  and  the  other  stand- 
ing committees  shall  include  at  least  one  member  of  the  Council. 

(e)  Sec.  8:  The  Council  shall  appoint  a  General  Board  of 
Examiners  to  pass  upon  applications  for  membership  received 
from  persons  not  residing  within  the  territory  of  any  section. 

Section  9:  All  committees  shall  be  directly  responsible  to  the 
Council,  and  shall  act  under  its  direction.  The  Council  may  at 
any  time,  at  its  own  discretion,  remove  any  or  all  members  of  a 
committee,  and  thereupon  the  President  shall  forthwith  appoint 
others  as  hereinbefore  provided ;  in  the  failure  of  the  President 
duly  to  appoint  such  a  committee,  the  Council  may  make  the  ap- 
pointment. The  terms  of  the  members  of  all  standing  and  tem- 
porary committees  shall  terminate  at  the  time  of  the  first  Council 
meeting  of  the  new  administration  of  each  year.  In  case  of  fail- 
ure to  appoint  new  standing  committees  on  Finance,  on  Papers 
and  on  Editing  and  Publication,  the  retiring  committees  shall 
continue  to  act  until  their  successors  are  appointed. 

(f)  Sec.  g:  So  far  as  possible,  all  reports  of  committees  to 
the  Council  shall  be  in  writing  and  signed  by  all  the  members 
of  the  Committee,  or  an  explanation  shall  be  offered  by  the 
chairman  for  the  absence  of  any  signature.  If  only  an  oral 
report  of  committee  work  can  be  rendered,  the  chairman  or 
other  member  making  such  report  shall  state  if  the  subject  mat- 
ter has  been  submitted  to  the  other  members  of  the  committee. 
and  shall  offer  an  explanation  if  this  has  not  been  done. 

By-laws  are  printed  in  small  type. 


18  TRANSACTIONS    I.    E.    S. — PART   I 

Section  10:  The  Committee  on  Finance  shall  have  direct  su- 
pervision of  the  financial  affairs  of  the  Society,  and  shall  pre- 
sent to  the  Council  an  annual  report  on  its  financial  condition. 
It  shall  approve  all  bills  before  payment,  and  shall  make  recom- 
mendations to  the  Council  as  to  the  investment  of  moneys  and 
upon  all  specific  appropriations.  No  payments  other  than  routine 
office  expenses  shall  be  made  by  the  General  Secretary  or  Treas- 
urer, except  upon  the  authorization  of  the  Committee  on  Finance. 

Section  11:  The  Committee  on  Papers  shall  have  general  su- 
pervision of  all  papers  to  be  presented  before  the  Society,  and 
shall  have  the  duty  of  preparing  the  programs  of  general  meet- 
ings of  the  Society  and  procuring  papers  for  presentation  before 
such  meeting.  No  paper,  discussion,  communication  or  report 
shall  be  printed  in  the  Transactions  of  the  Society  or  elsewhere 
until  approved  by  the  committee. 

(g)  Sec.  ii  :  The  Committee  on  Papers  may  direct  the  Com- 
mittee on  Editing  and  Publication  to  make  such  revision  as  may 
be  considered  necessary  or  desirable,  of  papers  and  communica- 
tions offered  for  publication ;  in  case  of  such  revision  the  manu- 
script shall  be  returned  to  the  author  to  obtain  his  consent  thereto, 
and  should  such  consent  be  refused,  the  paper  or  communication 
shall  not  be  accepted  for  presentation  before  the  Society. 

(h)  Sec.  ii  :  The  acceptance  of  a  paper  or  communication  for 
presentation  before  the  Society  or  any  section  thereof  shall  not 
be  considered  a  guarantee  of  its  publication  in  the  Transactions. 

Section  12:  The  Committee  on  Editing  and  Publication  shall 
edit  all  discussions  of  papers  presented  before  the  Society  or  any 
section  thereof,  and  shall  decide  all  questions  of  detail  regarding 
the  publication  of  papers,  discussions  and  communications.  The 
Transactions  and  other  publications  of  the  Society  shall  be  in, 
direct  charge  of  this  committee. 

(i)  Sec.  12:  The  Committee  on  Editing  and  Publication  may, 
at  its  discretion,  abridge  discussions  for  printing.  The  Com- 
mittee shall  cancel  remarks  that  do  not  bear  directly  on  the 
subject  under  discussion,  or  deal  in  personalities  or  have  mani- 
festly a  purely  commercial  object. 

(j)    Sec.    12:      All    papers,    discussions    and    other    matter    in- 
tended for  publication  in  the  Transactions  shall,  so  far  as  pos- 
sible, be  revised  and  edited  in  manuscript  and  not  in  proof. 
By-laws  are  printed  in  small  type. 


TRANSACTIONS    I.    E.    S. — PART   I  19 

(k)  Sec.  12:  A  revised  report  of  any  member's  discussion 
on  any  paper  must  be  received  at  the  general  office  of  the  Society 
within  ten  days  after  it  has  been  mailed  to  the  member,  otherwise 
revision  shall  be  made  by  the  Editing  Committee. 

(1)  Sec.  12:  The  Transactions  of  the  Society  shall  be  issued 
monthly,  except  during  the  three  summer  months. 

Section  13:  Five  members  shall  constitute  a  quorum  of  the 
Council.  The  "Vote  of  the  Council"  shall  be  a  vote  of  the 
majority  of  the  members  present  and  forming  a  quorum,  except 
where  a  letter-ballot  is  prescribed,  when  the  "Vote  of  the  Coun- 
cil" shall  be  a  vote  of  the  majority  of  the  entire  membership  of 
the  Council. 

ARTICLE  VIII. 

MEETINGS. 

Section  1:  The  annual  meeting  of  the  Society  shall  be  held 
on  the  second  Friday  of  June  of  each  year  at  a  place  designated 
by  the  Council,  when  a  report  of  the  proceedings  of  the  Society 
for  the  past  fiscal  year  shall  be  presented  by  the  Council,  which 
report  shall  be  verified  by  a  majority  of  the  Council,  including 
the  President,  Treasurer  and  General  Secretary. 

Section  2:  An  annual  convention  of  the  Society  shall  be  held 
on  a  date  and  at  a  place  fixed  by  the  Council,  for  the  presentation 
and  discussion  of  professional  papers  and  subjects.  The  Presi- 
dent shall  deliver  a  presidential  address  at  this  meeting. 

Section  3:  Other  meetings  of  the  Society  as  a  body  may  be 
held  at  such  time  and  place  as  the  Council  shall  direct,  at  which 
no  business  affecting  the  organization  or  policy  of  the  Society 
shall  be  transacted.  Notice  of  all  such  meetings  shall  be  sent 
by  mail  or  otherwise  to  all  members  at  least  ten  days  in  advance 
of  a  meeting. 

ARTICLE  IX. 

SECTIONS. 

Section  1:  Sections  of  the  Society  may  be  authorized  in  any 
State  or  locality  where  the  membership  exceeds  50. 

(a)  Sec.  I  :  Upon  petition  for  the  authorization  of  a  section 
of  the  Society,  the  Council  may  accord'  such  authorization  if  the 
necessary  membership  exists  within  the  locality  specified  in  the 
petition. 


By-laws  are  printed  in  small  type. 


20  TRANSACTIONS    I.    E.    S. — PART    I 

(b)  Sec.  i  :  Meetings  of  sections  shall  be  held  at  times  and 
places  fixed  by  the  Board  of  Managers.  When  suitable  papers 
or  lectures  are  available,  meetings  may  be  held  preferably  monthly 
except  during  the  three  summer  months. 

(c)  Sec.  i :  The  meetings  of  the  sections  shall  be  held  prefer- 
ably before  the  15th  of  the  month. 

Section  2:  Each  section  shall  nominate  and  elect  a  Chair- 
man, five  Managers  and  a  Secretary. 

Section  3:  The  officers  of  a  section  shall  be  elected  annually 
by  the  members  affiliated  with  the  section,  the  election  to  be  in 
accordance  with  a  procedure  fixed  by  the  Council. 

(d)  Sec.  3:  Procedure  in  nominating  and  electing  section  offi- 
cers shall  be  as  follows,  except  when  other  procedure  shall  be 
authorized  by  the  Council : 

A  section  nominating  committee  shall  be  appointed  by  the 
Section  Board  of  Managers  at  a  meeting  held  not  later  than 
March  1  of  each  year.  The  appointment  shall  be  reported  to 
the  General  Secretary.  This  committee  shall  consist  of  five 
members  of  whom  at  least  two  shall  be  past  officers  of  the 
section  or  members  of  the  Council. 

Not  later  than  March  15  of  each  year,  the  General  Secretary 
shall  notify  the  chairman  of  the  committee  that  it  is  the  com- 
mittee's duty  to  prepare  a  nomination  ticket  containing  the 
names  of  those  whom  they  deem  best  suited  for  the  section 
offices  to  be  filled  at  the  ensuing  annual  election.  The  report 
of  the  committee  shall  be  prepared  in  duplicate,  one  copy  shall 
be  submitted  to  the  chairman  of  the  section  and  the  other  copy 
shall  be  delivered  to  the  General  Secretary  not  later  than  April 
25.  The  ticket  thus  prepared  by  the  committee  on  nomination 
shall  be  printed  and  forwarded  to  all  section  members  not  later 
than  May  5,  in  connection  with  the  ballots  for  election  of  general 
officers.  The  election  of  section  officers  in  other  respects  shall 
be  carried  out  in  a  manner  similar  to  that  prescribed  for  the 
election  of  general  officers,  save  that  a  copy  of  the  report  of  the 
Committee  of  Tellers  on  the  results  of  the  section  election 
shall  be  mailed  as  soon  as  prepared,  to  the  chairman  of  the 
section  and  to  the  Chairman-elect. 

Section  4:  The  business  of  a  section  shall  be  conducted  by  a 
Board  of  Managers,  which  shall  consist  of  the  Vice-president  of 
the  Society  representing  the  locality  of  the  section,  and  the  Chair- 
man, Managers  and  Secretary  of  the  section. 

By-laws  are  printed  in  small  type. 


TRANSACTIONS    I.    E.    S. — PART    I  21 

Section  5:  The  Section  Board  of  Managers  shall  annually, 
at  the  first  meeting  of  the  society  year,  appoint  a  Board  of  Ex- 
aminers to  pass  upon  applications  for  membership. 

(e)  Sec.  5:  The  Board  of  Examiners  of  a  section  shall  con- 
sist of  the  Chairman,  the  Secretary  and  one  Manager  of  the 
section. 

Section  6:  A  section  may  formulate  by-laws  for  its  con- 
duct, which  shall  conform  with  the  Constitution  and  By-Laws 
of  the  Society  and  with  the  policy  of  the  Society  as  fixed  by  the 
Council.  Upon  approval  by  the  Council,  proposed  By-Laws 
may  be  adopted  by  a  two-thirds  vote  at  a  regular  or  special  meet- 
ing of  the  section ;  notification  of  such  meeting,  together  with  a 
copy  of  the  proposed  by-laws  shall  be  sent  to  all  members  of  the 
section  at  least  ten  days  prior  to  the  date  fixed  for  its  holding. 

Section  7:  Any  proposed  action  of  a  section  not  relating  to 
the  holding  of  meetings  and  the  discussion  of  papers  shall  be 
submitted  to  the  Council  of  the  Society  for  approval  prior  to 
being  put  into  execution. 

Section  8:  The  expenses  of  sections  incurred  for  postal-card 
notices  of  meetings  shall  be  paid  from  the  general  fund  of  the 
Society.  In  cases  where  there  is  no  desirable  auditorium  avail- 
able free  of  charge,  the  Council  shall  authorize  the  rental  of  a 
hall,  the  expense  to  be  payable  from  the  general  fund  of  the  So- 
ciety. Other  expenses  than  these  to  be  payable  from  the  gen- 
eral fund  of  the  Society  must  first  be  authorized  by  the  Council 
of  the  Society. 

(f)  Sec.  8:  The  Treasurer  may  deposit  with  the  Secretaries 
of  sections  a  sum  of  money,  the  amount  to  be  fixed  by  the 
Council,  to  provide  for  current  expenses. 

(g)  Sec.  8:  The  General  Secretary  of  the  Society  shall  supply 
to  each  section  all  stationery  and  printing,  aside  from  postal-card 
notices  necessary  for  the  conduct  of  its  business. 

Section  9:  A  Section  Board  of  Managers  may  authorize,  and 
shall  provide  for  the  payment  by  local  assessment  of  any  ex- 
penses of  a  section  beyond  those  authorized  to  be  paid  from  the 
general  fund  of  the  Society. 

By-laws  are  printed  in  small  type. 


22  TRANSACTIONS    I.    E.    S. — PART  I 

Section  10:  Papers  shall  be  approved  by  the  Section  Board 
of  Managers  prior  to  presentation  before  a  section.  Manu- 
script of  papers  approved  should  be  forwarded  to  the 
Committee  on  Papers  sufficiently  in  advance  of  date  of  presen- 
tation to  enable  advance  copies,  if  a  paper  be  approved  by  that 
committee  for  general  presentation,  to  be  printed  and  sent  to  all 
sections  for  distribution  prior  to  presentation  before  the  sections. 

Section  11:  Reports  of  discussions  shall  be  forwarded 
promptly  to  the  General  Secretary  who  shall  mail  them  at  once 
to  members  for  revision. 

(h)  Sec.  II:  The  Secretaries  of  sections  shall,  after  each 
meeting,  send  to  the  General  Secretary  a  statement  of  the  at- 
tendance and  of  the  business  transacted. 

(i)  Sec.  ii  :  The  Secretary  of  each  sction  shall  forward  to 
the  General  Secretary,  not  later  than  five  days  after  a  meeting  of 
a  section,  the  proceedings  of  the  meeting  for  publication  in  the 
Transactions. 

(j)  Sec.  ii  :  The  Secretaries  of  sections  shall  send  monthly 
to  the  General  Secretary  an  account  of  all  expenditures  in  the 
preceding  month. 

Section  12:  Should  the  membership  of  a  section  fall  below  50, 
or  the  average  attendance  at  meetings  not  warrant  the  expense 
of  maintaining  the  organization,  the  Council  may  cancel  its 
authorization. 

Section  13:  Sections  shall  abide  by  the  Constitution  and  By- 
Laws  of  the  Society  and  conform  to  the  regulations  of  the  Coun- 
cil. The  conduct  of  sections  shall  always  be  in  conformity  with 
the  general  policy  of  the  Society  as  fixed  by  the  Council. 

ARTICLE  X. 

LOCAIv  REPRESENTATIVES. 

Section  1:  When  authorized  by  the  Council,  the  President 
shall  appoint,  subject  to  the  approval  of  the  Council,  local  secre- 
taries or  local  committees  resident  in  cities  or  localities  where  it 
may  be  deemed  desirable  to  provide  representation  with  a  view 
to  promoting  the  work  of  the  Society. 

By-laws  are  printed  in  small  type. 


TRANSACTIONS    I.    E.    S. — PART   I  23 

(a)  Sec.  i :  Local  secretaries  shall  communicate  to  the  Gen- 
eral Secretary,  information  concerning  local  developments  in 
which  the  Society  may  be  concerned;  shall  endeavor  to  promote 
occasional  meetings  under  the  joint  auspices  of  the  Illuminating 
Engineering  Society  and  local  organizations  with  a  view  to 
fostering  interest  in  the  work  of  the  Society  and  shall  in  any 
other  manner  which  may  commend  itself  seek  to  develop  local 
knowledge  concerning  the  objects  of  the  Society  and  to  advise 
the  General  Secretary  when  opportunities  arise  for  the  Society 
to  promote  its  objects. 

(b)  Sec.  I :  Local  secretaries  may  obtain  Society  stationery 
upon  application  to  the  General  Office.  Local  secretaries'  expenses 
for  the  correspondence  may  be  billed  to  the  Society. 

ARTICLE  XL 

GENERAL. 

Section  1:  The  fiscal  year  of  the  Society  shall  be  October  i 
to  September  30. 

Section  2:  A  quorum  of  the  Society  shall  consist  in  number 
of  one-tenth  of  the  total  number  of  members  as  listed  in  the 
Society's  records  at  the  close  of  the  last  fiscal  year. 

ARTICLE  XII. 

AMENDMENTS    AND    BY-LAWS. 

Section  1:  Proposals  to  amend  this  Constitution  shall  be 
made  in  writing  to  the  Council  and  signed  by  at  least 
100  members  and  shall  reach  the  General  Secretary  not 
later  than  April  1.  The  Council  shall  consider  such 
proposals  and  direct  the  General  Secretary  to  send  out 
a  letter-ballot  on  their  adoption.  Votes  to  be  considered 
shall  be  received  not  later  than  May  26,  and  shall  be  re- 
ferred unopened  to  the  Committee  of  Tellers  who  shall  count 
such  votes  and  make  a  sealed  report,  which  shall  be  presented 
at  the  annual  meeting.  An  affirmative  vote  of  two-thirds  of  the 
entire  vote  cast  by  qualified  members  of  the  Society  shall  be 
necessary  to  secure  the  adoption  of  an  amendment.  An  amend- 
ment shall  take  effect  twenty  days  after  its  adoption. 

By-laws  are  printed  in  small  type. 


24  TRANSACTIONS     I.    E.    S. — PART    I 

Section  2:  By-Laws  in  interpretation  of  the  spirit  and  letter 
of  this  Constitution  and  for  its  execution  may  be  adopted  by  a 
majority  vote  of  the  entire  Council.  Votes  on  by-laws  shall  be 
by  letter  ballot.  Each  by-law  proposed  or  adopted  shall  state 
the  article  and  section  of  article  of  the  Constitution  to  which 
it  relates. 

(a)  Sec.  2 :  A  proposed  by-law  shall  not  be  acted  upon  at 
the  same  meeting  of  the  Council  at  which  it  is  submitted.  At 
least  ten  days  before  the  Council  meeting  at  which  a  by-law 
will  come  up  for  definite  action,  a  copy  of  the  same  shall  be 
forwarded  to  each  member  of  the  Council. 


By-laws  are  printed  in  small  type. 


TRANSACTIONS 


OF  THE 


Illuminating 
Engineering  Society 


JUNE,  1913 


PART  II 


K 


V 


Papers,  Discussions  and  Reports 


[  JUNE,  1913  ] 

CONTENTS  --  PART  II 


Some  Home  Experiments  in  Illumination  from  Large  Area 
Light  Sources.     By  Herbert  E.  Ives 229 

Gaslighting  in  an  Exhibition  Hall.     By  Robert  F.  Pierce. •  263 

Metal  Reflectors  for  Industrial  Lighting.     By  Thomas  W. 
Rolph 268 

Vision  as  Influenced  by  the  Brightness  of  Surroundings. 
By  Percy  W.  Cobb 292 

A  Practical  Solution  of  the   Problem  of  Heterochromatic 
Photometry.     By  Ch.  Fabry 302 


*> 


°l 


SOME    HOME    EXPERIMENTS     IN     ILLUMINATION 
FROM  LARGE  AREA  LIGHT  SOURCES. 


BY    HERBERT    E.    IVES. 


Synopsis:  The  experiments  here  described  were  carried  on  with  the 
consideration  of  efficiency  subordinated.  The  problem  was  to  obtain  a 
pleasing  satisfactory  illumination,  regardless  of  cost  or  of  established 
ideas.  Chief  attention  was  paid  to  direction,  diffusion  and  absence  of 
glare.  Means  of  measuring  these  qualities  not  being  established,  the 
criterion  adopted  was  the  writer's  judgment.  An  analysis  of  light 
sources  is  made  on  the  basis  of  size,  intrinsic  brilliancy,  direction  of  light 
and  whether  the  principal  light  source  is  visible  or  concealed.  A  study 
of  a  case  of  satisfactory  daylighting  from  windows  shows  the  window 
to  be  essentially  a  large  area  concealed  light  source  at  the  side.  A  number 
of  experimental  artificial  windows  are  described,  leading  to  an  estimate 
of  the  cost  of  an  exact  copy  of  daylight. 

INTRODUCTION. 

There  is  an  old  saying  that  "shoemakers'  children  go  barefoot," 
and  it  is  a  matter  of  common  observation  that  illuminating  en- 
gineers have  little  time  to  study  the  lighting  of  their  own  homes. 
This  should  not  be,  of  course,  for  a  "doctor  should  have  faith 
in  his  own  medicine."  But,  over  and  above  such  considerations, 
is  the  fact  that  the  least  cultivated,  but  probably  most  important, 
field  for  good  work  in  illumination  is  the  lighting  of  the  home. 
It  is  an  almost  everyday  experience  for  us  to  see  houses  which 
are  beautiful  by  day,  but  at  night  are  actual  atrocities — monu- 
ments to  a  lack  of  the  most  elementary  conception  of  the  lighting 
problem  on  the  part  of  the  architect. 

It  was  in  the  effort  to  light  my  own  home  satisfactorily  that  the 
experiments  to  be  described  were  begun.  As  time  went  on,  how- 
ever, the  problems  studied  became  somewhat  varied,  until  lately 
attempts  to  copy  daylight  distributions  of  illumination  have  occu- 
pied most  of  his  attention.  The  matter  which  follows  is  some- 
what rambling,  and  the  experiments  somewhat  incomplete,  for 
reasons  given ;  but  as  the  next  stage  of  the  work  may  be  de- 
layed for  some  time  it  was  thought  well  to  publish  this  much,  if 
for  no  other  reason  than  for  the  discussion  which  may  result. 

*  A  paper  presented  at  a  meeting  of  the  Philadelphia  section  of  the  Illuminating 
Engineering  Society,  May  16,  1913. 


23O  TRANSACTIONS  I.  E.  S. — PART  II 

Out  of  this  study  have  come  some  novel  lighting  devices  and 
some  new  lighting  schemes  which  are  thought  worth  recording. 
Also,  in  the  course  of  the  work  brightness  measurements  of  dif- 
ferent conditions  were  made  which  bring  out  points  of  interest. 

ADVANTAGES  AND   DISADVANTAGES   OF 
HOME  EXPERIMENTS. 

A  great  advantage  of  using  one's  own  house  as  an  illumina- 
tion laboratory  is  that  one  can  live  with  the  lighting  system  and 
learn  how  it  wears,  as  the  various  normal  activities  are  carried 
on  at  their  normal  times.  The  home  rooms  have  their  furniture 
in  place,  which  is  a  tremendous  advantage.  Everyone  who  has 
moved  to  a  new  house  knows  how  very  different  a  room  appears 
before  and  after  furnishing.  It  is  the  difference  between  dis- 
comfort and  comfort,  and  to  attempt  to  judge  a  lighting  scheme 
in  a  bare  laboratory  room  means  too  serious  a  handicap.  Often 
a  mere  touch  of  decorative  treatment  will  reclaim  a  lighting 
device  which,  in  its  crude  form,  is  hard  to  imagine  desirable. 

The  disadvantages  are  several.  A  considerable  one  is  that  it 
is  difficult  and  often  impossible  to  make  changes  in  the  positions 
of  the  outlets.  A  certain  amount  of  flexibility  is  achieved  by 
using  base-board  connections,  but  after  a  more  or  less  unsightly 
structure  of  wires  or  tubes  has  arisen,  the  time  is  likely  to  come 
when  something  logically  next  on  the  program  cannot  be  tried 
out  without  rebuilding  the  house.  This  limits  experimentation, 
consequently  some  of  the  things  described  in  this  paper  are  in- 
complete, and  the  descriptions  must  end  with  suggestions  as  to 
what  might  be  better.  The  usual  separation  from  proper  tools 
and  the  time  taken  to  make  what  should  be  trivial  changes  consti- 
tute other  disadvantages.  A  special  house  with  furniture  com- 
plete would  possibly  be  the  right  thing  for  a  study  of  this  kind,  but 
even  this  would  labor  under  the  disadvantage  that  one  would  be 
likely  not  to  feel  "at  home,"  and  would  not  be  apt  to  live  in  it 
long  enough  to  determine  how  it  wore. 

A  CHARACTERISTIC  OF  HOME  LIGHTING. 
A  characteristic  of  home  lighting  which  makes  it  well  worth 
studying  by  a  lighting  company  is  that  efficiency  counts  for  little. 
Furniture    and    pictures    are    not    bought    for    their    cheapness, 


IVES:     SOME    HOME   EXPERIMENTS   IN    ILLUMINATION         23I 

neither  should  light  be — nor  will  it  when  the  public  becomes  bet- 
ter educated  in  the  use  of  light  and  in  its  remarkable  decorative 
possibilities.  In  domestic  lighting  the  real  problem  is  to  obtain 
a  pleasant  illumination  effect,  almost  without  regard  to  cost.  In 
the  experiments  here  described  no  attention  whatever  has  been 
paid  to  efficiency.  The  author  has  put  himself  in  the  position  of 
a  householder  who  demands  certain  lighting  effects  and  is  willing 
to  pay  for  satisfaction.  In  fact,  I  believe  that  many  possible 
lighting  schemes  have  not  been  tried  at  all — notably  copies  of 
daylight  distribution — because  the  illuminating  engineer  has  been 
apt  to  think  too  quickly  of  efficiency.  He  has  been  an  engineer 
before  being  an  artist,  whereas  the  more  fruitful  procedure  is  to 
first  obtain  the  desired  effect  and  then  count  the  cost.  It  may 
happen  that  the  price  of  the  paint  prevents  the  painting  of  a  mas- 
terpiece. Even  if  it  costs  too  much  now  to  have  just  what  we 
would  like,  we  must  look  forward  to  the  day  of  more  efficient 
light  sources.  We  may  even  bring  that  day  nearer  by  finding 
greater  needs  for  them. 

MEASUREMENTS  AND  MEASURING  INSTRUMENTS. 

This  paper  is  to  be  comparatively  free  from  foot-candle  values, 
watts  per  square  foot  and  the  like.  The  chief  points  aimed  at 
are  diffusion,  proper  direction,  freedom  from  excessive  contrast 
and  glare,  a  pleasant  non-fatiguing  quality  and  good  appearance. 
How  can  these  be  measured?  In  the  writer's  opinion  the  only  re- 
liable and  sensitive  instrument  is  the  experienced  eye  of  one  who 
has  observed  and  thought  about  illumination  for  a  period  of 
years.  We  have  all  of  us  noticed  how  sensitive  we  have  become 
to  exposed  light  sources.  Formerly  they  irritated  us,  to  be  sure, 
but  we  did  not  know  the  cause.  Now  we  choose  our  position  with 
care,  shield  our  eyes  with  our  hands,  and  in  other  ways  give 
testimony  to  the  vastly  increased  sensitiveness  of  our  seeing 
mechanism.  Meanwhile  it  remains  a  sad  fact  that  we  have  gen- 
erally available  no  method  of  measuring  lighting  conditions 
which  will  do  more  than  distinguish  between  two  so  extreme  in 
quality  that  a  mere  casual  glance  will  tell  which  is  good  and 
which  is  bad.  Until  we  are  better  off  in  this  respect  the  trained 
eye  must  be  encouraged  as  the  best  we  have.  It  is  worse  than 
3 


232  TRANSACTIONS  I.  E.  S. — PART  II 

useless  to  make  measurements  of  factors  which  are  not  the  ones 
really  vital.  If  home  illumination  demands  diffusion  and  a  feel- 
ing of  comfort  meter-candles  illumination  on  the  working  plane 
has  very  little  to  do  with  the  problem.  It  must  be  borne  in  mind 
that  measurements  are  of  no  value  unless  they  record  conditions 
whose  qualities  are  proved  by  experience.  Experience  can  never 
be  dispensed  with.  It  must  be  called  upon  sooner  or  later.  This 
paper  deals  with  lighting  schemes  whose  measuring  instrument 
at  present  is  experience.  Recognizing  this  fact  I  have  out- 
spokenly given  here  as  the  final  criticism  my  own  judgment  of 
the  success  of  an  installation.  In  all  cases  however,  I  have  had 
as  many  comments  and  suggestions  from  others  as  I  could  obtain. 

REMARKS  ON  DIRECTION,  DIFFUSION,  GLARE 
AND  CONTRAST. 

One  peculiarity  of  daylight  illumination  from  windows  is  its 
direction.  Most  artificial  systems  cast  the  light  downward  from 
a  point  near  the  center  of  the  ceiling.  There  then  occurs  in 
changing  from  day  to  artificial  light  a  90  degree  rotation  of  shad- 
ows. By  day  they  are  long  and  sweep  across  the  room ;  by  night 
they  are  short  or  completely  covered  by  the  shadow-casting 
object.  A  clear  illustration  of  this  change  is  to  be  seen  in  the 
ordinary  railway  car.  By  day  the  illumination  comes  from  the 
windows  at  the  two  sides ;  at  night  from  the  high  centrally  placed 
units.  As  far  as  direction  is  concerned,  side  wall  brackets,  floor 
standards  or  table  lamps  on  side  tables  approximate  more  nearly 
to  window  conditions. 

Diffusion  of  light  with  soft  shadows,  as  is  well  known,  is  most 
perfectly  brought  about  by  the  use  of  a  large  area  of  light  source 
and  is  one  of  the  chief  merits  of  the  "indirect"  system.  A  cer- 
tain degree  of  diffusion  may  also  be  obtained  by  a  multiplicity  of 
light  sources,  although  this  is  apt  to  give  merely  a  multiplicity  of 
sharply  defined  shadows,  instead  of  the  soft  shadows  of  the  large 
source.  The  ordinary  window  with  its  area  of  5  or  6  square 
feet  (0.46  or  0.56  sq.  m.),  gives  a  far  better  degree  of  diffusion 
than  most  artificial  schemes.  Another  factor  in  diffusion  is  the 
color  of  the  light.  Blue  light  diffuses  better  than  yellow.  It  is 
therefore  quite  possible  that  daylight  is  inherently  apt  to  be  better 
diffused  than  is  yellow  artificial  light. 


IVES:     SOME   HOME  EXPERIMENTS  IN  ILLUMINATION        233 

In  regard  to  contrast  and  glare,  it  is  well  to  bear  in  mind  some 
numerical  data.  A  Welsbach  mantle  has  an  intrinsic  brilliancy 
of  35  candle-power  per  square  inch  (6.45  sq.  cm.)  ;  a  tungsten 
filament  about  1,000  and  a  patch  of  sky  about  2  to  3.  For  the 
same  intensity  of  illumination,  that  is,  the  same  general  bright- 
ness of  illuminated  objects,  therefore,  the  physical  contrast  be- 
tween the  brightness  of  the  illuminated  objects  and  the  light 
sources  themselves  will  be  from  ten  times  to  hundreds  of  times 
greater  with  visible  "direct"  artificial  light  than  with  daylight. 
If  the  light  sources  are  concealed,  this  contrast  may  be  avoided, 
but  it  occurs  again  to  a  disturbing  degree  if  specularly  reflecting 
surfaces  are  present.  These  reflect  images  of  the  light  sources 
with  about  1/10  or  V20  °f  tne  intrinsic  brightness  of  the  parent 
source.  It  is,  in  fact,  immediately  evident  to  an  expert  whether 
the  lighting  of  a  room  is  due  to  small  or  large  sources  by  an 
inspection  of  these  two  things — the  sharpness  of  the  shadows  and 
the  brightness  of  the  specular  reflections.  Half  of  this  handicap 
of  small  sources  is  eliminated  by  entirely  avoiding  specularly 
reflecting  surfaces. 

SOME  CHARACTERISTICS  OF  DAYLIGHT  OUT-OF-DOORS. 

Out-of-doors  daylight  may  be  pleasant  or  unpleasant.  In  a 
previous  paper*  which  records  brightness  measurements  of 
typical  outside  illuminations,  certain  conditions  were  found  to  be 
most  pleasing.  These  were  an  excess  of  brightness  in  the  upper 
hemisphere,  usually  with  a  maximum  near  the  horizon,  and  (a 
very  essential  condition)  sunlight  directed  from  the  side  to  give 
long  shadows. 

A  criticism  which  has  been  made  of  many  "indirect"  installa- 
tions is  that  the  bright  "sky"  furnished  by  the  light  ceiling  is 
bounded  by  dark  walls  carried  up  above  the  eye  line.  This  gives 
an  effect  of  being  down  in  a  well,  or  as  though  the  room  had 
been  lifted  up  away  from  one.  To  produce  a  real  out-of-doors 
daylight  distribution,  the  above  quoted  work  would  call  for  an 
extension  of  this  brightness  down  to  the  horizontal.  This  would 
tend  to  overcome  an  objection  frequently  made  to  the  'indirect" 
system — that  faces  are  subject  to  unnatural  downward  shadows. 

*  The  Distribution  of  Luminosity  in  Nature,  Trans.,  I.  E.  S.,  p.  6S7,  Vol.  VI  (1911). 


234 


TRANSACTIONS  I.  E.  S. — PART  II 


Out-of-doors  our  faces  receive  considerable  horizontal  illumina- 
tion. In  addition  the  habit  of  wearing  hats  must  be  taken  into 
account.  Hats  protect  the  eyes  from  overhead  bright  areas  and 
also  reduce  the  vertical  component  of  the  illumination  on  the 
face. 

The  effect  of  the  large  sky  angle  on  the  extreme  brightness 
ratio  must  not  be  overlooked.  Out-of-doors  the  sky  approxi- 
mates to  an  infinite  plane.  Under  this  a  white  surface  will  be  as 
bright  as  the  illuminant  while  other  objects  will  be  bright  in 
proportion  to  their  reflecting  powers.  As  the  area  of  "sky"  is 
decreased,  its  brightness  remains  the  same,  but  that  of  illuminated 
objects  becomes  less,  thereby  increasing  the  extreme  brightness 
ratio  between  unconcealed  illuminant  and  illuminated  objects. 
Diffusion  is  also  decreased  and  shadows  take  on  a  more  definite 
downward  direction. 

AN  ATTEMPT  TO  COPY  OUT-OF-DOORS  DIFFUSED  DAYLIGHT. 
The  guiding  idea  here  was  to  carry  the  direct  illumination  of 
an  "indirect"  system  down  on  to  the  walls  in  such  manner  as  to 
approximate  the  out-of-doors  conditions,  and  thereby  produce 
a   pleasant   diffused   illumination.     The    room   available    was    a 


Fig  i. — Diffused  illumination  from  side  walls. 

small  one,  about  14  feet  (4.27  m.)  square,  furnished  with  a  cen- 
tral fixture  and  papered  a  medium  light  buff.  Under  this  central 
fixture  was  suspended  a  yellowish  Japanese  umbrella  of  oiled 
paper,  convex  side  up.  The  direct  light  of  the  lamps  thus  fell 
on  the  walls,  which  acted  as  the  chief  light  source  of  the  room, 
since  the  transmission  of  the  umbrella  as  used  was  low,  while  the 


IVES:     SOME   HOME  EXPERIMENTS   IN   ILLUMINATION         235 

ceiling  was  little  brighter  than  the  walls  and  largely  invisible  be- 
cause of  the  comparatively  small  size  of  the  room.     (See  Fig.  1.) 

A  peculiar  merit  of  this  scheme  lay  in  the  splendid  illumination 
of  the  pictures  on  the  walls.  These  had  to  be  slightly  tilted  to 
prevent  reflection  of  the  bright  light  sources,  and  in  the  direction 
of  the  entrance  doorway  a  translucent  Japanese  fan  was  placed 
above  the  umbrella  to  conceal  the  lights.  This  installa- 
tion was  used  for  many  months  and  gave  great  satisfaction. 
The  "well"  appearance  sometimes  noted  in  indirect  systems  was 
entirely  absent,  There  was  no  tendency  for  the  eye  to  wander 
upward  to  a  bright  point  above.  Although  the  illumination 
seemed  low,  it  proved  ample  for  sewing  and  other  close  work, 
which  was  done  without  the  fatigue  customary  to  working  with 
"artificial"  light.  A  pleasing  addition  to  the  general  diffused 
lighting  was  furnished  by  placing  a  table  lamp  on  one  side, 
whereby  the  direct  light  of  the  sun  was  copied.  One  point 
proved  to  my  satisfaction  by  this  experiment  is  that  light  walls, 
well  illuminated,  are  not  productive  of  evil.  On  the  contrary, 
they  help  in  a  marked  degree  the  general  effect  of  brightness.  A 
room  looks  brighter  with  bright  walls  and,  unless  some  extreme 
case  is  taken  this  means  that  the  object — to  light  the  room — has 
been  attained. 

This  lighting  arrangement  was  only  given  up  when  I  moved 
away  from  Cleveland.  It  needs  a  small  room,  or,  in  a  large  room, 
a  number  of  ceiling  outlets  not  far  from  the  walls  in  place  of  one 
central  one.  It  is  intended  at  some  future  time  to  place  outlets  in 
a  large  room  which  will  make  possible  lighting  the  walls  in  a 
similar  manner  and  the  expectation  is  that  it  will  furnish  a  very 
satisfactory  illumination.  In  some  indirect  or  semi-indirect  in- 
stallations it  might  be  of  interest  to  tilt  the  units  near  the  walls 
in  order  to  secure  a  similar  effect.  These  units  might  be  made 
purposely  unsymmetrical. 

CLASSIFICATION   OF   LIGHT   SOURCES. 

The  common  classification  into  direct,  indirect  and  semi-in- 
direct, while  convenient,  is  by  no  means  a  complete  analysis.  The 
true  classification  is  on  the  basis  of  the  size  of  the  light  source 
and  whether  it  is  concealed  or  visible.     The  earlier  light  sources, 


236  TRANSACTIONS  I.  E.  S. — PART  II 

such  as  the  gas  flame  and  the  carbon  filament  were  small  sources 
of  relatively  high  intrinsic  brilliancy.  From  them  we  have  been 
working  toward  large  light  sources  of  lower  intrinsic  brilliancy. 
The  fact  that  some  of  these  are  large  areas  of  diffusely  reflecting 
ceiling  and  that  others  are  large  areas  of  diffusely  transmitting 
glassware  is  of  no  great  significance.  It  is  of  significance  that 
the  light  source  is  of  large  extent  and  of  low  intrinsic  brilliancy. 
It  is  also  of  significance  that  in  each  of  these  cases  ("indirect" 
and  "semi-direct")  the  principal  light  source  is  visible.  By 
"principal  light  source"  I  mean  the  source  which  furnishes  the 
greater  part  of  the  light  which  falls  on  floor,  furniture  and  work- 
ing plane.  It  is  also  of  great  importance  that  as  ordinarily 
worked  out,  in  "indirect"  and  "semi-indirect"  installation,  the 
light  source  is  of  necessity  the  brightest  object  in  the  field  of 
view.  Therefore,  the  indirect  and  the  semi-indirect  systems  sift 
down  to  this :  that  the  unconcealed  light  source  is  made  as  large 
as  is  possible  in  the  attempt  to  make  its  intrinsic  brilliancy  low 
and  bearable.  Incidentally  this  causes  great  diffusion  of  light 
and  soft  shadows. 

In  some  forms  of  "direct"  lighting  another  device  is  employed, 
namely,  partial  or  complete  concealment  of  the  light  source.  In 
this  case  it  may  result  that  the  principal  light  source  is  not  the 
brightest  object  in  the  field  of  view.  If  the  concealment  is  suffi- 
cient there  may  be  a  complete  absence  of  points  of  high  in- 
trinsic brilliancy — more  so  than  in  the  "indirect"  system — except 
for  points  of  specular  reflection  of  the  light  sources  if  polished 
surfaces  are  present.  These  latter  are  of  high  intrinsic  bril- 
liancy in  the  "direct"  systems  and  low  in  the  "indirect." 

It  will  be  noted  by  those  who  have  followed  this  analysis  that 
while  we  have  lighting  systems  with  large,  visible,  low-brightness 
sources,  and  systems  with  small,  either  visible  or  invisible,  high 
brightness  sources,  there  are  none  consisting  of  large,  low  bright- 
ness concealed  light  sources — (with  the  possible  exception  of 
some  forms  of  deck  lighting  where  the  light  is  carried  well  above 
a  transparent  or  only  partly  diffusing  glass).  As  will  be  pointed 
out  later,  this  is  the  specification  (when  that  of  direction  from 
the  side  is  included)  of  pleasant  daylight  illumination  from  win- 
dows. 


IVES:     SOME   HOME   EXPERIMENTS    IN    ILLUMINATION         237 

A  LARGE  OVERHEAD  LIGHT  SOURCE. 
It  follows  from  the  classification  just  given  that  the  charac- 
teristics of  "indirect"  lighting  would  be  obtained  without  the 
process  of  reflection  if  a  large  enough  low-brilliancy  source  was 
used.  This  was  experimentally  accomplished  in  a  very  simple 
manner  by  the  use  of  a  yellowish  oiled  paper  Japanese  umbrella 
of  4  feet  diameter,  turned  point  down,  within  which  are  the  lights 
in  frosted  globes.  (See  Fig.  2.)  A  certain  amount  of  light 
falls  on  the  ceiling,  so  that  it  might  be  called  a  "semi-indirect" 
system,  but  by  far  the  greater  part  of  the  illumination  is  due  to 
the  visible  light  source.  This  is  of  rather  low  intrinsic  brilliancy, 
so  that  it  can  be  looked  at  without  any  sensation  of  strain,  and 
gives  ample  light  all  over  the  16  foot  by  20  foot  room  with  an 
emission  of  about  7,500  lumens. 


Fig.  2. — Illumination  from  large  area,  low  brightness  overhead  source. 

Now  as  to  its  merits  and  defects.  Because  of  the  size  and  low 
intrinsic  brilliancy  shadows  are  soft  and  specular  reflections  are 
too  dull  to  be  annoying.  Because  of  the  conical  shape  a  better 
illumination  of  the  walls  and  far  corners  is  obtained  than  would 
be  possible  by  a  ceiling  reflection  scheme.  On  the  other  hand, 
the  direction  of  the  shadows  is,  for  the  center  of  the  room  at 
least,  straight  down  and,  therefore,  different  from  the  most 
agreeable  daylight  condition.  But  the  greatest  defect  lies  in  the 
fact  that  the  principal  light  source  is  visible,  and,  low  as  is  the 


238  TRANSACTIONS  I.  E.  S. — PART  II 

intrinsic  brightness,  it  is  the  brightest  object  visible  and  is  too 
bright  to  be  continuously  in  the  field  of  vision.  When  the  occu- 
pants of  the  room  sit  around  the  center  table  or  sit  sideways  to 
the  light,  the  illumination  is  extremely  satisfactory.  If,  how- 
ever, when  a  number  of  people  are  conversing,  they  face  each 
other  across  the  center,  then  in  time  the  large  bright  umbrella 
becomes  an  irritant — far  less,  of  course,  than  would  the  usual 
bright  points,  but  still  noticeable  to  a  sensitive  and  critical  eye. 

An  experiment  was  tried  with  a  still  larger  light  source,  but 
it  had  the  same  defects.  The  conclusion  was  arrived  at  that  by 
no  process  of  increasing  the  size  of  the  source,  unless  it  actually 
occupied  the  whole  upper  hemisphere — which  is  out-of-doors 
daylight  on  a  sunless  day — could  the  light  source  be  made  in- 
nocuous. Diffusion  is  not  alone  sufficient.  Any  system  by  which 
the  light  source  is  the  brightest  visible  thing  gives  an  extreme 
brightness  ratio  which  is  too  great  to  wear  well.  If,  in  the 
room  illuminated  with  the  large  umbrella,  the  hand  is  held  over 
the  eyes,  or  an  eye  shade  is  worn,  it  leaves  little  to  be  desired ; 
there  remains  merely  the  question  of  the  direction  of  the  light 
to  be  objected  to;  indeed  many  might  prefer  the  centrally  located 
source.  Of  course  the  higher  the  light  source  the  less  this  de- 
fect would  be,  so  that  the  "indirect"  practise  of  making  the  ceil- 
ing the  real  source  chooses  the  best  condition,  for  it  then  approxi- 
mates, in  a  small  room,  to  a  concealed  source. 

What  seems  to  be  most  desirable  in  this  particular  room 
is  some  means  of  illumination  with  this  large  low  brightness 
source,  at  the  same  time  keeping  it  concealed.  That  done,  the 
range  of  visible  brightness  is  not  too  great.  As  will  be  seen 
below,  daylight  illumination  by  windows  does  something  of  this 
sort. 

CHARACTERISTICS  OF  PLEASANT  DAYLIGHT  ILLUMINATION 
FROM    WINDOWS. 

I  have  no  brief  for  daylight  illumination  in  general,  nor  for 
all  kinds  of  window  lighting.  Excessively  unpleasant  illumina- 
tion may  be  found  by  day.  On  the  other  hand,  daylight  at  its 
best  has  never,  in  my  opinion,  had  a  rival  for  general  excel- 
lence.    There  has   been   much   discussion   on   this   point,   some 


IVES':     SOME  HOME  EXPERIMENTS  IN   ILLUMINATION         239 

declaring  flatfoot  for  daylight  as  necessarily  best,  others  record- 
ing themselves  as  preferring  the  "mellow,"  "cozy"  artificial  light. 
As  a  matter  of  fact,  I  believe  no  one  has  ever  had  a  good  copy 
of  daylight  to  use  at  night  in  place  of  the  ordinary  artificial 
systems.  Perhaps  it  is  a  case  of  "sour  grapes"  with  those  who 
say  they  would  not  copy  daylight  if  they  could.  I  believe,  too, 
that  illuminating  engineers  have  neglected  analysis  of  the  good 
and  bad  points  of  daylight.  Without  such  analysis  it  is  futile  to 
hope  to  copy  the  merits  or  avoid  the  defects. 

My  observation,  common  to  that  of  others,  is  that  window 
lighting  is  not  always  satisfactory.  It  may  be  insufficient,  or  it 
may  cause  excessive  contrast  in  brightness.  In  my  own  house 
the  first  floor  living  room  has  one  door  and  one  window  on  the 
front  and  two  windows  at  the  side  facing  another  house.  Ordi- 
narily the  room  is  not  sufficiently  lighted  by  daylight.  The  front 
window  is  too  narrow.  If  the  shade  is  run  up  to  the  top  it  is  im- 
possible to  look  with  comfort  at  the  window,  because  the  large 
patch  of  bright  sky  then  visible  is  in  violent  contrast  to  the  unil- 
luminated  wall  at  the  sides  of  the  window.  Even  with  the 
shades  down,  so  that  the  sky  is  not  seen,  the  houses  opposite  are 
apt  to  present  too  great  contrast  of  brightness  to  the  walls  adja- 
cent to  the  window.  Only  under  one  condition  is  the  daylight 
really  satisfactory  and  that  is  when  a  spot  of  sunlight  falls  well 
back  on  the  floor,  the  houses  opposite  being  in  the  shade.  Then 
the  walls  adjacent  to  the  window  receive  light  from  the  floor,  the 
ceiling  receives  light  from  the  floor  and  from  the  pavement  out- 
side and  all  is  bright  and  cheerful.  With  snow  on  the  ground 
outside,  the  effect  is  still  better. 

Another  case  of  bad  daylight  illumination  is  furnished  in  the 
upper  floors  of  a  tall  office  building  which  is  not  faced  by  any 
other  structure.  If  one  sits  facing  a  window,  one  sees  a  large 
expanse  of  bright  sky,  which  soon  becomes  painful.  In  these 
rooms  the  official  habitually  places  his  desk  so  that  his  back  is 
toward  the  windows.  Visitors  who  face  the  windows  do  not  find 
daylight  pleasant. 

Most  fortunately  a  case  of  the  other  kind  is  to  be  found  in  my 
own  house,  namely,  an  instance  of  a  pleasant  and  satisfactory 
daylight  illumination.     This   is  in  a  second  floor  room,   facing 


240 


TRANSACTIONS  I.  E.  S. — PART  II 


south,  illuminated  by  two  adjacent  windows  in  a  bay  nearly  the 
width  of  the  room.  The  wall  paper  is  a  figured  gray  of  approxi- 
mately 45  per  cent,  reflection  coefficient.  The  dimensions  of  the 
room  are  shown  in  Fig.  3.  A  long  room  of  this  type  is  perhaps 
one  of  the  hardest  to  illuminate  satisfactorily  by  daylight  or  arti- 
ficial daylight  because  of  its  depth.  One  with  windows  on  two 
sides  would  have  been  preferable  and  easier  to  deal  with,  if  it 
had  been  available. 


/o 


:.*> 


Fig.  3. — Distribution  of  surface  brightness  in  arbitrary  units  by  daylight.    Large  values, 
a  case  where  sunlight  falls  on  floor.    Small  values,  sun  not  streaming  in. 


Observations,  both  qualitative  and  quantitative,  made  from 
time  to  time  in  this  room,  revealed  some  interesting  points.  ( 1 ) 
At  all  times,  raising  the  shades  above  the  middle  sashes  made 
the  lighting  less  pleasant.  (2)  The  room  was  brightest  and,  on 
the  whole,  most  pleasing  when  the  sun  streamed  in  on  the  floor 
or  wall  and  shone  on  the  street  below.  (3)  The  effect  was 
unpleasant  when  the  sun  fell  upon  the  (gray  stone)  house  fronts 
opposite.  (4)  A  gray  or  misty  day  was  not  pleasant.  (5)  The 
distribution  of  illumination  or  brightness  was  such  that  the  face 
of  a  person  sitting  with  his  back  or  side  to  the  window  received 
sufficient  light  from  the  room  so  as  not  to  be  in  marked  contrast 
to  the  background,  or  to  have  the  two  sides  of  his  face  unduly 
different  in  brightness. 

Two  sets  of  measurements  of  brightness  are  plotted  in 
Fig.  3,  one  for  the  case  of  sunlight  streaming  in  the  window,  the 


IVES:     SOME   HOME   EXPERIMENTS   IN   ILLUMINATION         24I 

other  when  it  is  on  the  street  below,  but  not  streaming  in  nor  on 
the  houses  opposite.     Both  effects  were  extremely  satisfactory. 

An  analysis  of  the  various  data  has  led  me  to  the  following 
conclusions : 

The  most  pleasant  conditions  are  those  when  the  effective  light 
source  is  of  three  parts :  ( i )  A  large  area  of  bright  sky  sub- 
tending the  solid  angle  of  the  window,  invisible  from  the  greater 
part  of  the  room  and  illuminating  the  floor  and  lower  part  of  the 
room.  (2)  A  large  area  of  intrinsic  brilliancy  about  one-tenth 
that  of  the  sky,  visible  from  the  room  (houses  opposite).  (3)  A 
large  area  of  intrinsic  brilliancy  about  one-fourth  to  one-fifth  that 
of  the  sky,  invisible  from  the  greater  part  of  the  room,  but  illu- 
minating its  ceiling. 

Add  that  these  light  sources  are  at  the  side,  and  remember 
that  the  bright  sky  is  of  much  lower  intrinsic  brilliancy  than  most 
artificial  light  sources,  and  the  complete  specification  of  pleasant 
window  daylight  of  this  room  becomes  something  like  this: 
illumination  from  large  area  concealed  light  sources  at  the  side, 
that  is,  just  the  case  we  have  seen  to  be  absent  from  prevalent 
lighting  systems. 

There  must  be  added  to  this  set  of  conditions  the  further  one 
that  the  walls  adjacent  to  the  window  must  be  well  lighted  in 
order  to  prevent  excessive  contrast.  This  is  frequently  accom- 
plished by  having  windows  on  two  sides.  In  the  present  case  the 
walls  adjacent  to  the  windows  are  perpendicular  to  these  and 
are,  therefore,  lighted  by  one  set  of  windows. 

Probably  to  an  architect,  who  has  made  a  study  of  window 
lighting,  this  is  all  an  old  story.  It  is  interesting  in  the  light  of 
these  observations  to  notice  that  low  broad  windows  (frequently 
in  alcoves)  are  becoming  common  in  the  newer  houses,  taking  the 
place  of  the  old-fashioned  high  narrow  window,  the  upper  por- 
tion of  which  sometimes  cannot  be  used  without  exposing  to  view 
a  dazzling  patch  of  sky. 

THE  WINDOW  AS  A  LIGHT  SOURCE. 

The  numerical  data  given  in  Fig.  3  make  possible  a  study  of 
the  windows  of  this  particular  room  as  a  light  source. 

As  viewed  from  a  point  in  the  room  the  windows  present  prac- 


242 


TRANSACTIONS  I.  E.  S. — PART  II 


tically  the  appearance  of  two  adjacent  squares  30  in.  by  30  in. 
(0.762  m.  x  0762  m.),  with  their  centers  at  a  height  of  42  in. 
(1.07  m.)  above  the  floor.  As  viewed  from  various  points  the 
angle  subtended  by  the  light  source  of  course  changes,  diminishing 
to  zero  above  and  below  and  to  either  side.  Were  the  landscape  as 
seen  through  the  window  of  uniform  intrinsic  brilliancy,  then  the 
window  would  be  equivalent  to  a  flat  uniformly  bright  plate.     It 


,^«S* 


U4j>AJ 


Fig.  4.— The  window  considered  as  a  point  source  of  light. 

could  be  copied  by  a  sheet  of  thick  opal  glass  illuminated  from 
outside,  or  by  a  diffusely  reflecting  surface,  such  as  an  "indirect" 
ceiling  source  turned  through  900.  This,  however,  is  not  the 
case,  except  when  in  a  high  building  facing  the  sky,  or  if  a  fog 
reduces  all  out-of-doors  to  a  nearly  uniform  gray  (a  ratio  of 
sky  to  houses  of  2  to  1  was  found  by  measurement  on  a  foggy 


IVES:     SOME   HOME  EXPERIMENTS   IN   ILLUMINATION         243 

day).  Both  of  these  conditions  are  found  to  be  undesirable,  as 
noted  above. 

In  the  second-story  room,  which  was  studied  chiefly,  the 
change  of  intrinsic  brilliancy  and  effective  area  may  be  followed 
best  by  consulting  the  drawing,  Fig.  4,  in  which  the  window  is 
considered,  for  convenience  only,  as  a  point  source  of  light.  Start- 
ing directly  below  the  window  the  illumination  is  from  a  narrow 
slit  of  the  intrinsic  brilliancy  of  the  sky.  As  the  test  surface  is 
moved  away  from  the  window  this  slit  increases  in  width  (as  the 
cosine  of  the  angle  with  the  normal)  until  a  point  (A)  well  back 
on  the  floor  is  reached.  Here  the  top  of  the  houses  opposite  be- 
comes visible,  cutting  off  part  of  the  sky.  At  the  point  (B)  the 
sky  is  no  longer  visible  and  from  there  to  the  point  (C)  high  up 
on  the  wall  the  intrinsic  brilliancy  of  the  window  is  that  of  the 
house  opposite.  Between  (C)  and  (D)  the  street  below  becomes 
visible  and  from  (D)  over  the  ceiling  to  the  window  the  illumina- 
tion is  from  a  patch  of  the  brightness  of  the  street  surface.  There 
are.  therefore,  to  be  distinguished  (in  this  case)  three  different 
brightnesses,  the  distribution  of  illumination  from  each  of  which 
is  represented  by  a  circle  of  appropriate  diameter,  tangent  to  the 
window.  From  the  brightness  measurement  of  Fig.  3,  it  ap- 
pears that  these  three  circles  should  have  diameters  for  the  most 
pleasant  condition  of  about  10-1-3,  and  they  are  so  represented 
in  the  figures. 

It  will  be  seen  that  a  seated  person  in  almost  any  part  of  the 
room,  looking  toward  the  window,  sees  only  the  comparatively 
dull  houses  opposite.  The  floor  and  ceiling  are  illuminated  by  a 
much  brighter  source  than  the  houses  and  act  as  secondary  light 
sources  of  low  brightness.  If  the  sun  strikes  the  floor  the  spot 
becomes  another  secondary  light  source.  The  most  startling 
thing  is  the  very  irregular  shape  of  the  light  distribution  curve 
and  its  sharp  transition  from  maximum  to  minimum,  unequalled, 
I  believe,  in  any  commercial  lighting  unit.  Were  it  not  for  the 
fact  that  the  source  (window)  is  large  these  sharp  transitions 
would  cause  sharp  contrasts  of  light  and  shadow  between  the 
portion  of  wall  and  ceiling  represented  at  A,  B,  C,  etc.  Actually, 
the  deviation  of  the  window  from  a  point  source  makes  the 
transitions  gradual  and  soft. 


244  TRANSACTIONS   I.  E.  S. — PART  II 

An  interesting  feature  of  the  distribution  from  a  window  is 
that  it  tends  to  diminish  the  extreme  brightness  ratio.  The  great- 
est illumination  is  on  the  floor  and  furniture,  which  are  almost 
invariably  the  darkest  surfaces.  With  a  uniform  distribution 
from  a  side  window  or  with  many  artificial  systems  the  floor 
and  furniture  retain  the  normal  brightnesses  which  their  low 
reflecting  powers  give  them.  With  windows  they  are  made  sev- 
eral fold  brighter,  which  must  be  no  inconsiderable  factor  in  the 
general  effect.  This  fact  explains  in  part  why  with  "indirect" 
systems  the  floor  and  furniture  are  apt  to  appear  unnaturally 
dark.  In  order  to  preserve  the  daylight  ratio  of  brightness  above 
and  below  it  would  be  necessary  with  the  "indirect"  system,  with 
its  bright  ceiling,  to  exchange  the  floor  coverings  and  furniture 
for  some  with  at  least  a  4  or  5  fold  higher  coefficient  of  reflec- 
tion. The  excellent  appearance  of  white  tiled  barber  shops  un- 
der the  "indirect"  system  is  confirmatory  of  this. 

ATTEMPTS  TO  ARTIFICIALLY  APPROXIMATE  THE  CHOSEN 
DAYLIGHT   DISTRIBUTION. 

The  General  Problem. — As  an  interesting  exercise,  attention 
was  turned  to  ways  and  means  for  reproducing  artificially  the 
best  daylight  conditions  in  the  room  described.  It  soon  became 
evident  that  the  problem  was  not  a  simple  one.  It  appears,  in 
fact,  that  there  is  just  one  solution,  and  that  is  to  reproduce  en- 
tire the  broad  expanse  of  sky,  houses  and  street.  No  small  light 
source  will  copy  all  the  qualities  desired.  For  instance,  while  it 
is  possible  to  produce  at  a  given  point  in  the  room  the  general 
distribution  of  illumination  and  brightness  given  by  daylight,  the 
same  lighting  scheme  will  not  produce  the  desired  effect  at  an- 
other point.  The  problem  is  closely  akin  to  that  of  producing 
a  perspective  drawing  which  would  automatically  change  its  per- 
spective as  the  observer  moved,  or  a  photograph  in  which  one 
could  look  around  the  corners  by  moving  one's  head  from  side  to 
side.  For  instance,  a  horizontal  plate  placed  above  and  outside 
the  window  would  produce  the  desired  sky  effect  near  the  win- 
dow, but  would  fail  farther  back  in  the  room.  A  vertical  plate 
of  low  intrinsic  brilliancy  would  take  the  place  of  the  opposite 
houses  as  far  as  the  eye  of  an  observer  within  the  room,  but  the 


1VKS:     SOME    HOME   EXPERIMENTS   IN    ILLUMINATION"         245 

sweep  of  tenfold  brighter  light  across  the  floor  and  the  brighter 
light  on  the  ceiling  would  be  missing.  As  it  was  not  possible  to 
light  up  "all  out  of  doors,"  several  approximations  to  the  condi- 
tions given  in  Fig.  3  were  attempted,  which  are  described  below. 
The  first  of  these  may  be  called  a  lighting  fixture,  the  others 
lighting  schemes,  or,  as  a  non-technical  visitor  described  one  of 
them :  "not  lights,  but  light." 

(i)  A  Side  Wall  Fixture  to  Approximate  a  Window  Effect. — 
This  fixture  might  be  popularly  described  as  "semi-indirect  from 
the  side,"  although  it  presents  a  number  of  deviations  from  what 
would  be  obtained  by  merely  supporting  a  semi-indirect  bowl  out 
from  the  wall.  In  construction  it  might  be  most  nearly  described 
by  saying  it  is  a  table  lamp  with  the  back  half  and  top  of  the 
shade  removed.  In  detail  it  consists  of  the  light  source,  which 
should  be  of  such  form  or  provided  with  such  a  reflector  as  to 
throw  its  light  chiefly  below  and  above  the  horizontal  about  as 
does  the  prototype  window  curve,  added  to  which  is  a  large  area 
translucent  screen  so  calculated  as  to  let  the  direct  light  fall  upon 
the  wall  behind,  the  ceiling  above  and  the  whole  floor  of  the  room, 
and  of  such  degree  of  translucency  as  to  closely  approximate  the 
brightness  of  the  illuminated  wall  behind.  The  device  is  shown 
in  Fig.  5,  where  it  is  represented  as  a  floor  standard.  From  the 
plan  and  elevation  it  will  be  seen  that  for  no  normal  position  in 
the  room  is  the  bright  part  of  the  light  source  visible,  but  screen 
and  wall  form  a  large  source  of  low  intrinsic  brilliancy.  This  is 
secured  by  the  peculiar  shape  of  the  screen,  which  is  concave  to 
the  light  with  respect  to  the  horizontal  plane,  half  enclosing  it, 
but  only  large  enough  vertically  to  conceal  the  light  from  the 
occupants  of  the  room.  The  screens  so  far  made  up  have  been 
covered  with  flowered  silk,  or  cretonne,  and  lend  themselves  ad- 
mirably to  the  general  decorative  scheme. 

This  fixture  when  used  in  a  room  having  light  walls  (a 
necessary  condition)  gives  to  a  large  degree  the  desired  window 
effect.  The  light  is  directed  from  the  side,  giving  the  long  pleas- 
ing shadows;  there  is  a  large  measure  of  diffusion,  due  to  the 
large  area  of  the  effective  source ;  the  floor  and  furniture  have  a 
pleasant  "sun  lit"  appearance.  The  thing  of  beauty  is  the  light 
itself  and  not.  as  in  many  cases,  a  piece  of  decorative  metal  work. 


246 


TRANSACTIONS  I.  K.  S. — PART  II 


designed  to  be  viewed  by  daylight.  The  limitations  to  this  device 
are  the  necessity  for  bright  walls,  which  is  not  serious,  and  the 
fact  that  the  direct  light  in  the  lower  hemisphere  cannot  be  made 
to  fall  as  far  out  on  the  floor  as  with  windows,  without  carrying 
the  shade  too  far  in  front  of  the  light  for  convenience.     It  does, 


'         / 


Fig.  5. — A  side  wall  unit  to  approximate  window  conditions. 

however,  make  an  extremely  pleasing  light  source  and  has,  in 
several  months  of  use,  worn  well.  It  is  especially  good  if  two  or 
more  are  grouped  at  the  window  end  of  the  room.  It  appears 
that  the  best  height  for  these  shades,  if  they  are  to  be  in  the  field 
of  vision,  is  about  on  the  level  of  the  eyes.  If  a  light  source 
must  be  visible  it  seems  to  be  less  irritating  at  this  line  than  at 


IVES:     SOME    HOME   EXPERIMENTS   IN    ILLUMINATION         247 


some  degrees  above.  It  is  perfectly  possible  to  make  the  shades 
and  wall  behind  of  such  brightness  that  a  person's  face  may  be 
viewed  against  them  as  a  background  without  undue  contrast. 

Side  wall  brackets  have  always  been  popular  with  those  who 
wish  their  faces  to  appear  at  their  best.  The  defect  of  most 
bracket  fixtures  has  been  that  the  light  sources  have  an  altogether 
too  great  intrinsic  brilliancy  and  make  excessive  contrast  with 
the  wall  behind,  especially  if  the  latter  is  dark  wood,  as  it  too 
often  is.  This  defect  is  overcome  in  the  arrangement  here  de- 
scribed. There  has  thus  far  been  no  dissenting  opinion  as  to  the 
excellent  performance  of  these  side  wall  "windows." 


0+ 


Fig.  6.— A  point  source  copy  of  daylighting.     (Brightness  values  are 
multiplied  by  a  factor  10). 

(2-)  An  Alcove  Light  with  Semi-window  Characteristics. — 
This  scheme  was  tried  out  after  a  study  of  the  window  consid- 
ered as  a  point  source  as  shown  in  Fig.  4,  and  was  carried 
through  in  full  knowledge  of  the  fact  that  the  characteristics  of 
a  large  window  could  not  be  completely  copied  without  a  com- 
paratively large  source.  The  plan  is  shown  in  Fig.  6.  where 
"A"  represents  a  line  of  light  sources  with  aluminized  scoop 
reflectors  sending  a  large  part  of  their  light  in  the  lower  hemi- 
sphere and  approximating  the  lower  portion  of  the  large  circle  of 
Fig.  4.  At  "B"  is  placed  a  frame  15  in.  x  6l/2  ft.  covered  with 
cretonne  and  muslin  sheeting.  This  to  an  observer  back  in  the 
room  takes  the  place  of  the  houses  opposite,  as  a  large  area  low 
brightness  source.  By  trial  such  a  thickness  of  fabric  was  found 
4 


248  TRANSACTIONS  I.  E.  S. — PART  II 

that  the  illumination  back  in  the  room  due  to  it  was  one-tenth 
that  from  the  row  of  lights  as  seen  below  it  from  the  floor,  that 
is,  the  condition  found  best  by  daylight. 

This  installation  is  exceedingly  interesting.  With  7,200 
lumens  the  surface  brightnesses  through  the  room  measure  up  on 
an  average  one-twelfth  that  found  as  the  average  of  the  two 
daylight  distributions  given  in  Fig.  3.  It  was  found  most  desir- 
able, however,  to  let  considerable  light  go  to  the  ceiling,  which 
gives  more  light  in  the  room  and  approximates  more  to  the  "sun- 
light" case  of  daylight  distribution  in  Fig.  3.  This,  when  the  eye 
is  adapted  to  night  conditions,  is  very  nearly  ample  at  every  point 
of  the  room.  Probably  twice  as  much  would  be  more  than 
enough.  The  room  presents  a  strikingly  daylight  appearance, 
except  for  three  things,  all  due  to  the  same  cause.  First,  the  line 
of  demarcation  between  the  light  from  the  'houses  opposite"  and 
the  "sky"  is  too  sharply  marked  on  the  walls ;  second,  the  specu- 
lar reflections  on  the  furniture,  books,  etc.,  near  the  windows, 
visible  on  entering  the  room,  are  relatively  far  brighter  than  they 
would  be  by  day,  as  is  to  be  expected ;  third,  the  shadows  of  the 
furniture  are  too  sharp.  These  are  all  due  to  the  point  source 
character  of  the  light,  and  are  only  to  be  seen  above  and  below 
the  parts  illuminated  by  the  cloth  covered  screen.  The  appear- 
ance is  practically  the  same  as  can  be  obtained  by  covering  all 
the  windows  except  a  narrow  horizontal  slit.  Wtihin  the  angle 
which  receives  light  only  from  the  screen  and  from  walls  and 
ceilings,  the  daylight  character  is  almost  startling.* 

(3)  A  Window  Reflector. — The  requirement  of  a  large-area 
concealed  light  source  at  the  side,  having  the  distribution  curve 
of  the  window  as  a  light  source,  may  be  approximated  to  by 
combinations  of  optical  devices,  such  as  prisms  and  reflecting 
surfaces.  One  of  these  devices  forms  the  last  experiment  to  be 
described. 

In  accordance  with  the  subject  matter  of  this  paper,  it  is  clear 
that  an  "indirect"  fixture,  if  directed  toward  the  wall,  would  not 

*  If  the  windows  be  made  horizontal  slits  ol  about  10  inches  width  and  if  the  time  of 
day  is  taken  when  the  sun  falls  on  the  houses  opposite,  an  extreme  case  of  bad  lighting 
is  obtained.  The  illumination  of  the  room  is  quite  insufficient;  there  is  a  high  ratio  of 
brightness  between  the  light  source  and  its  surrounding  in  the  field  of  view;  shadows  are 
sharp  and  specular  reflections  are  too  bright.  This  illustrates  clearly  the  importance  of 
the  light  source  being  large  and  of  having  its  brightest  part  concealed. 


IVES:     SOME   HOME  EXPERIMENTS  IN   ILLUMINATION        249 

give  the  desired  window  effect  because,  while  the  light  source 
would  be  large,  it  would  be  of  uniform  brightness  as  viewed  from 
all  directions.  In  order  to  secure  for  this  experiment  the  proper 
unsymmetrical  distribution  of  brightness,  recourse  was  had  to  re- 
flecting surfaces  intermediate  between  specular  and  matt.  Meas- 
urement of  a  number  of  surfaces,  such  as  scratch-brushed  alumi- 
num, aluminum  paint,  mirror  glass  covered  with  transparent  cur- 
tain material,  etc.,  led  to  the  choice  of  ribbed  mirror  glass,  sand- 
blasted on  the  front.  Later  it  was  found  better  to  decrease  the 
brightness  of  the  diffuse  component,  and  this  was  done  by  rub- 
bing the  sandblasted  surface  slightly  with  oil.  Probably  a  lighter 
sandblasting  would  have  given  the  desired  quality.     Such  mir- 


Fig.  7. — Section  of  special  reflectors  used  for  artificial  window. 

rors  placed  with  the  ribs  vertical  and  illuminated  by  a  light  in 
front  and  above  give  a  large  semi-specularly  reflected  component, 
which  is  nevertheless  so  much  diffused  as  to  appear  to  come  from 
a  large  area  of  mirror;  and  a  small  diffuse  component,  the  ratio 
between  the  normal  and  reflected  brightness  being  easily  made 
one-tenth  by  tilting  the  mirror. 

A  pair  of  frames  of  these  was  made,  as  shown  in  Fig.  7,  each 
frame  consisting  of  four  7  in.  by  30  in.  elements.  These  frames 
were  hung  in  the  window,  as  shown  in  Fig.  8,  and  were  illumi- 
nated by  a  row  of  small  units  in  the  aluminized  concentrating 
reflectors  used  before.  As  shown  in  Fig.  7,  the  specular  reflection 
from  the  three  upper  elements  illuminates  the  floors  and  lower 


250 


TRANSACTIONS  I.  E.  S. — PART  II 


parts  of  the  walls  of  the  room.  The  bottom  elements  illuminate 
the  ceiling,  taking  the  place  of  the  brightly  lit  street  pavement 
before. 

An  observer  sitting  in  the  room  sees  these  surfaces  as  of  a  uni- 
form low  brightness.  On  dropping  his  head  he  begins  to  see  the 
top  elements  brighten,  corresponding  to  the  sky  appearing  over 
the  house  tops.  On  dropping  still  lower,  the  other  elements 
brighten  until  at  the  floor  the  whole  "window"  appears  a  uniform 
sheet  of  light.  On  standing  up  and  approaching  the  window  the 
lowest  elements  become  bright  in  the  same  manner.  This  copy 
of  window  conditions  is  in  fact  very  close.  The  reflected  light 
is,  because  of  the  ribs  and  the  sandblasting,  excellently  diffused. 
Shadows  in  the  room  are  long  and  soft. 


! 


Fig.  8. — Artificial  window  lighting.     (Brightness  values  multiplied  by  factor  10). 

The  excellence  of  the  copy  is  shown  by  the  relative  brightness 
measurements.  These  indicate  that  the  relative  brightness  of  all 
visible  objects  is  nearly  the  same  as  by  daylight  without  the  sun 
on  the  floor,  as  shown  in  Fig  3,  the  actual  brightness  being  close 
to  one-tenth  of  that  of  daylight.  In  fact  this  imitation  of  day- 
light distribution  in  the  room  is,  from  the  standpoint  of  relative 
brightness,  direction  of  light  and  diffusion,  so  perfect  as  to  leave 
outstanding  only  the  factors  of  actual  intensity  of  illumination 
and  color  of  light. 

Good  and  Bad  Features. — This  reflector  scheme  has  many  ele- 
ments of  merit.  It  is  in  some  respects  almost  uncanny,  espe- 
cially to  one  not  looking  toward  the  imitation  windows.  The 
sweep  of  light  over  the  floor;  the  illumination  of  the  low  por- 


IVES:     SOME   HOME   EXPERIMENTS  IN   ILLUMINATION        251 

tion  of  the  walls;  the  brightness  ratio  between  ceiling  and  floor 
are  most  striking  and  very  different  from  the  usual  artificial  light 
conditions.  But  the  general  effect  is  not  pleasing.  There  is  a 
harsh  "contrasty"  effect.  The  darker  end  of  the  room  looks  too 
dark ;  the  windows  top  light ;  shadows  are  too  inky  black. 
The  effect  is  very  much  like  a  foggy  day,  or  as  though  the  houses 
opposite  were  brilliantly  sunlit.  Why  is  this?  In  some  part  to 
mental  bias,  perhaps;  the  windows  do  not  look  just  like  real  win- 
dows and  one  is  strongly  conscious  of  the  fact.  Also  the  fact 
that  the  light  source  is  near,  instead  of  distant,  and  that  our  ocu- 
lar muscles  of  accommodation  are  quite  conscious  of  the  real 
state  of  affairs.  The  explanation  of  the  greater  part  probably 
lies  in  two  facts :  first,  the  absolute  intensity  is  too  low,  and,  sec- 
ond, the  color  of  the  light  is  not  white.  The  absolute  intensity  is 
about  one-tenth  daylight  intensity.  My  observation  leads  me  to 
believe  that  a  given  physical  ratio  of  brightness  becomes  sub- 
jectively greatly  increased  when  the  order  of  illumination  is  de- 
creased as  much  as  it  is  here.  If  a  ratio  of  brightness  of  10-1 
for  different  points  in  the  room  is  satisfactory  by  daylight  then 
with  1 -10  that  illumination  this  ratio  must  be  decreased  to  5-1 
or  3-1.  With  this  low  order  of  illumination  the  shadows,  such 
as  those  on  the  room  side  of  the  face,  are  so  dark  as  to  appear 
nearly  black;  the  high-light  side  alone  is  seen.  By  the  greater 
light  of  day,  although  the  relative  brightness  of  the  two  sides  of 
the  face  is  the  same,  the  dark  side  is  bright  enough  to  be  easily 
seen ;  the  bright  side  does  not  appear  too  bright.  I  can  best  de- 
scribe the  phenomenon  by  saying  that  the  room  under  this  1-10 
illumination  looks  like  an  under-exposed  photograph — the  high 
lights  alone  come  out.  By  exposing  longer  a  good  negative  is 
obtained;  by  increasing  the  illumination  tenfold,  I  believe — un- 
less the  color  difference  has  a  great  deal  to  say — that  this  copy  of 
daylight  would  be  "the  real  thing."  Physically  it  measures  up 
in  the  right  proportion.  Subjectively  the  proportion  becomes 
distorted.  I  have  found  support  for  this  idea  by  the  experiment 
of  wearing  dark  glasses  of  1-10  transmission  in  this  room  when 
the  daylight  conditions  are  good.  The  conditions  become  trying 
at  once.  What  is  called  for  in  the  present  case  is  an  increase  in 
total  light,  which  cannot  be  made  in  the  existing:  house  installa- 


252  TRANSACTIONS    I.  E.  S. — PART    II 

tion  without  danger.  Incidentally  it  may  be  remarked  that  this 
accentuation  of  contrasts  merely  means  a  larger  value  for  the 
Fechner  fraction  at  low  illuminations,  which  is  actually  found  by 
experiment. 

The  general  effect  in  the  room  is  enormously  improved  by  add- 
ing the  light  from  two  side  wall  table  lamps  described  above, 
placed  in  the  corners  next  the  alcove.  These  give  somewhat 
more  light,  but  more  than  that,  they  throw  their  added  light  upon 
the  ceiling  and  side  walls  without  increasing  the  brightness  of  the 
"windows."  These  latter  cease  to  be  too  bright ;  the  shaded  side 
of  the  face  becomes  visible  and  a  first  rate  effect  is  produced. 

On  turning  off  the  window  light  alone  the  room  seems  to  jump 
upward,  showing  very  prettily  the  importance  of  the  daylight 
effect  of  the  sweep  of  light  on  the  lower  part  of  the  room.  I  be- 
lieve this  window  experiment  would  have  been  more  successful 
had  the  room  been  provided  with  an  artificial  window  at  the  side. 
A  long  dark  room  such  as  this  is  difficult  to  light  by  day ;  were  it 
not  on  the  south  side  of  the  house  I  believe  it  might  be  a  failure. 
It  certainly  represents  an  extreme  type.  I  am  inclined  to  think 
another  type  of  room  say  a  broad  shallow  one  furnished  with 
these  artificial  windows  might  be  quite  successful,  even  without 
an  increase  in  the  light. 

The  color  of  the  light  may  have  something  to  do  with  the  harsh 
effect.  As  above  noted  blue  light  scatters  much  better  than  yel- 
low. A  photograph  by  ultra-violet  light  will  show  almost  no 
shadows  owing  to  scattered  light  from  the  dust  particles  in  the 
air,  while  in  infra-red  light  shadows  are  black.  Something  of 
this  sort  may  contribute  to  the  inadequacy  of  this  artificial 
window. 

SUMMARY  OF  CONCLUSIONS  AND  SPECULATIONS  ON  ROOM 
LIGHTING  BY  LARGE  AREA  SOURCES. 

Several  conclusions  which  seem  to  be  justified  by  these  various 
experiments,  are  given  as  follows : 

The  desirable  qualities  sometimes  found  in  window  lighting  by 
daylight  are,  (1)  direction  from  the  side;  (2)  soft  shadows  and 
low  intensity  of  specular  reflection,  due  to  the  large  size  of  the 
source;  (3)  the  direction  of  a  large  part  of  the  light  on  the  lower, 


IVES:     SOME   HOME   EXPERIMENTS   IN   ILLUMINATION         253 

usually  the  darker,  parts  of  the  room,  and  (4)  the  concealment  of 
the  principal  light  source. 

It  appears  possible  to  make  a  very  close  physical  copy  of  the 
window  as  a  light  source  whereby  all  the  characteristics  may  be 
duplicated.  An  attempt  to  do  this  shows  that  to  the  above  quali- 
ties must  be  added  (5)  large  quantity  of  light  as  compared  with 
what  is  usually  available  from  artificial  sources.  It  is  suggested 
that  this  is  less  necessary  if  the  artificial  windows  are  placed  so 
as  to  secure  a  more  uniform  distribution  of  illumination  than  is 
necessary  by  daylight. 

These  various  qualities  can  be  separated  to  some  extent  into 
necessities  and  luxuries;  some  are  necessary  for  comfort,  others 
appeal  to  the  esthetic  sense.  I  believe  that  in  concealment  of  the 
light  source,  so  that  it  is  not  the  brightest  object  visible,  in  making 
it  of  large  area  and  in  introducing  a  certain  amount  of  side  light 
from  large  area,  very  low  intrinsic  brilliancy  sources,  most  of  the 
really  necessary  and  many  of  the  attractive  characteristics  of 
daylight  may  be  obtained,  with  much  lower  intensity  and,  conse- 
quently, lower  cost  than  a  true  copy  of  daylight  would  come  to. 
I  favor  a  combination  of  deck  lighting  for  general  floor  and 
working  plane  illumination,  with  side  wall  lights  of  the  type  de- 
scribed above.  I  hope  to  report  on  an  installation  of  this 
type  at  some  future  date.  If  the  room  is  not  too  large  nothing  at 
present  promises  to  excel  the  side  wall  window  brackets  of  Fig.  5. 

THE  COST  OF  COPYING  DAYLIGHT. 

If  it  were  not  for  the  prohibitive  cost  of  an  exact  copy  of  good 
window  daylight,  it  would  be  a  very  desirable  thing.  This  paper 
may  be  concluded  with  a  calculation  of  just  what  this  cost  is. 

The  room  experimented  with  has  an  area  of  200  square  feet, 
1,600  lumens  were  used  and  ten  times  that  were  required,  or 
16,000.  Suppose  absorbing  screens  were  used  to  make  this  light 
of  daylight  color ;  this  would  call  for  about  ten  times  the  amount 
of  light,  or  160,000  lumens  generated.  Taking  the  tungsten  lamp 
at  1.5  watts  per  spherical  candle  as  the  light  source,  one  watt  per 
square  foot  is  actually  used,  ten  watts  is'  called  for,  one  hundred 
if  subtractive  daylight  is  made  (or  20  kilowatts  for  the  room). 

Let  us  finish,  however,  with  a  little  speculation  and  prophecy. 


254  TRANSACTIONS   I.  E.  S. — PART    II 

I  have  shown  elsewhere*  that  if  we  could  produce  white  light 
with  no  accompanying  invisible  radiation  or  other  losses,  by  some 
merger  of-  fire-flies,  there  should  result  an  efficiency  of  330  lumens 
per  watt,  or  about  40  times  that  of  the  tungsten  lamp.  Since 
16,000  lumens  is  necessary  to  produce  daylight  intensity  and  dis- 
tribution in  this  room,  there  would  be  required  for  this  ideal 

1  •      1-1      16,000  .  ,rt,  .     .    .    . 

white  light  ,  or  about  so  watts.    This,  it  is  interesting  to 

6  300  ° 

note,  is  just  about  what  the  builder  of  the  house  has  provided 
for  in  the  low-hung,  glaring  and  utterly  horrible  central  fixture. 
Artificial  daylight  may  not  always  be  a  luxury. 

DISCUSSION. 

Prof.  George  A.  HoadlEy  :  It  seems  to  me  that  Dr.  Ives 
has  given  us  in  the  paper  which  is  before  us  an  excellent  example 
of  how  to  carry  on  laboratory  experimentation  in  our  own  homes. 

It  is  true  that  the  subject  under  consideration  lends  itself 
more  than  some  others  would,  to  such  investigation  but  there 
are  many  phases  of  the  problems  of  lighting  that  require  just  the 
kind  of  semi-leisure  investigation  that  can  be  carried  out  at 
home  during  the  long  winter  evenings. 

Each  problem  that  comes  up  in  the  lighting  of  our  homes  is 
different  from  that  which  presents  itself  to  our  neighbor.  They 
all  have,  however,  one  common  property,  they  are  intensely 
practical  problems  and  if  successfully  solved  will  tend  to  secure 
greater  comfort  and  economy. 

I  fully  agree  with  Dr.  Ives  in  his  contention  that  large  light 
sources  are  the  most  desirable.  In  order  to  secure  such  a  source 
in  my  dining  room,  I  made  a  simple  change  in  the  fixture  that 
has  proved  to  be  most  satisfactory.  The  lamp  shade  was  of 
fluted  glass  dark  green  on  the  outside  and  white  on  the  inside, 
giving  a  good  downward  diffusion.  Its  location  was  too  far 
from  the  ceiling  and  the  light  from  the  lamp  shone  directly  in 
the  eyes  of  those  sitting  at  the  table.  By  the  insertion  of  a  ground 
glass  disk  in  the  shade  holder  below  a  globe-shaped  lamp,  the 
light  comes  from  the  area  of  a  circle  fourteen  inches  in  diameter, 
gives  uniform  diffusion  and  no  glare. 

*  Electrical  World,  June  15,  1911. 


SOME    HOME   EXPERIMENTS   IN    ILLUMINATION  255 

Dr.  C.  E.  Ferree  :  I  have  seen  the  devices  described  by  Dr. 
Ives,  and  I  agree  with  him,  so  far  as  my  present  knowledge  of 
the  subject  goes,  on  all  the  essential  points  of  his  discussion. 
The  table-lamp  shade  seems  to  me  to  be  especially  good.  It  is 
in  principle,  I  think,  the  best  device  of  its  kind  I  have  yet  seen. 
Among  other  features  it  possesses  the  advantage  of  shielding 
the  eye  from  the  source  of  light  without  interfering  with  the 
distribution  in  any  other  direction.  A  shade  of  this  kind  should 
not  only  give  excellent  results  for  desk  and  table  work  but  it 
can  be  used  very  successfully  for  the  illumination  of  a  room. 
Its  degree  of  opacity,  its  distance  from  the  light-source,  and  the 
distance  of  the  light-source  from  the  reflecting  walls  of  the 
room,  can  all  be  so  regulated  that  a  fairly  high  degree  of  uni- 
formity of  general  illumination  can  be  obtained, — very  much 
more  uniform  than  is  usually  obtained  from  fixtures  of  the 
semi-indirect  type.1  The  table-lamp  has,  moreover,  so  far  as 
the  eye  is  concerned,  an  advantage  of  position  over  the  ceiling 
fixture,  or  a  wall  fixture  at  the  height  at  which  wall  fixtures  are 
usually  placed,  because  it  can  be  kept  more  nearly  at  the  level 
of  the  eye.  That  is,  if  we  are  to  have  the  source  of  light  in  the 
field  of  vision  at  all,  it  is  better  to  have  it  as  nearly  as  possible 
at  the  level  of  the  eye,  for  the  image  of  the  source  when  it  falls 
on  the  retina  in  its  horizontal  meridian  produces  less  discomfort 
than  when  it  falls  in  the  vertical  meridian.  When  used  as  a  desk- 
light  I  think  the  shade  should  be  made  more  opaque  than  when 
used  for  the  general  illumination  of  a  room,  because  when  work- 
ing at  a  desk  the  lines  of  sight  are  directed  downwards  and  a 
shade  of  the  degree  of  transparency  used  by  Dr.  Ives  would 
permit  more  light  to  fall  on  the  sensitive  lower  half  of  the  retina 
than  should  be  the  case  if  the  maximum  degree  of  comfort  is 
to  be  attained.  With  a  shade  of  the  dimensions  used  by  him 
probably  the  most  favorable  distribution  of  light  over  the  retina 
can  be  secured  by  making  the  shade  completely  opaque.  The 
condition  to  be  attained  is  that  the  illumination  of  the  retina 
shall  fall  off  more  or  less  uniformly  from  center  to  periphery. 

1  The  classification  of  this  shade  as  semi-indirect  in  type  is  somewhat  arbitrary. 
Dr.  Ives  may  very  well  prefer  to  call  it  indirect  since  by  a  proper  regulation  of  the 
opacity  of  the  shade,  distance  from  the  wall,  etc..  the  light  in  the  immediately  surround- 
ing field  may  be  made  approximately  equal  in  intensity  to  that  at  the  source. 


256  TRANSACTIONS   I.  E.  S. — PART   II 

In  securing  this  condition  the  factors  that  must  be  taken  into 
account  and  regulated  in  this  case  are  obviously  opacity  and 
breadth  of  shade,  direction  and  distance  from  the  eye,  position 
in  room  relative  to  reflecting  walls,  etc.  The  shade  in  principle 
permits  and  in  fact  needs  more  or  less  special  regulation  for 
each  individual  case. 

I  found  the  light  from  the  inverted  umbrella  uncomfortable 
to  the  eye,  and  from  the  similarity  of  the  distribution  given  by 
it  to  the  distribution  given  by  systems  we  have  already  studied, 
I  would  conjecture  that  the  eye  would  also  fall  off  considerably 
in  efficiency  when  working  under  it.  With  regard  to  this  light, 
the  following  points  may  be  noted.  (1)  The  distribution  given 
by  it  is  semi-indirect  in  type.  In  our  work  on  distribution  we 
have  found  that  the  eye  falls  off  in  efficiency  almost  as  badly 
under  the  semi-indirect  systems  as  under  the  direct.  (2)  It  is  a 
ceiling  light  and  because  of  the  low  ceilings  found  in  most  dwell- 
ings it  would  have  to  be  placed  at  a  height  above  the  level  of  the 
eye  that  would  be  very  uncomfortable.  As  it  is  installed  in  Dr. 
Ives's  home,  its  angle  of  direction  from  the  eye  is  very  near  to 
that  which  we  have  found  to  give  the  greatest  discomfort.  (3) 
The  light  is  rendered  yellowish  by  transmission  through  the 
shade.  So  far  as  our  work  on  the  effect  of  quality  of  light  on 
the  eye  has  been  carried,  we  have  found  greater  loss  of  efficiency 
and  more  discomfort  under  yellow  light  than  under  lights  whiter 
in  quality.  (4)  Sufficient  diffusion  is  not  produced  by  the  um- 
brella as  a  shade  to  break  up  entirely  the  images  of  the  light- 
source. 

Dr.  Ives's  photometric  analysis  of  the  distribution  of  light  in 
a  room  illuminated  by  daylight  from  windows  is  interesting  and 
very  suggestive  as  to  methods  of  attack  on  the  problem  of  light- 
ing. His  device  to  reproduce  this  distribution  for  artificial  light 
I  find,  however,  to  be  very  uncomfortable  to  the  eye.  This 
difference  in  effect  on  the  comfort  of  the  eye  presents  an  inter- 
esting problem  for  solution.  Dr.  Ives  states  that  he  has  repro- 
duced by  means  of  his  artificial  window,  at  a  lower  scale  of 
intensity,  the  relative  distribution  of  light  in  the  room  gotten 
from  the  daylight  window.  That  is,  the  ratio  of  intensity  of 
light  at  the  source  to  that  at  various  points  in  the  surrounding 


SOME   HOME  EXPERIMENTS  IN   ILLUMINATION  257 

field  was  made  the  same  for  both  cases.  From  the  standpoint 
of  the  light  in  the  room,  then,  the  only  difference  between  the 
two  cases  is  in  terms  of  his  statement  apparently  in  the  quality 
of  light  and  in  the  scale  of  intensity  used.  This  difference,  I 
understand,  is  believed  by  Dr.  Ives  to  be  the  cause  of  the  differ- 
ence in  the  effect  on  the  comfort  of  the  eye.  I  should  not  myself 
be  inclined  fully  to  accept  this  explanation  until  more  differential 
evidence  is  obtained.  The  effect  of  both  factors  in  question  can 
be  investigated  under  conditions  in  which  it  is  more  definitely 
certain  that  no  other  factor  is  present.  For  example,  the  effect 
of  quality  of  light  on  the  comfort  of  the  eye  can  easily  be  tested 
out  by  separate  experiment,  also  the  effect  of  changing  the  scale 
of  intensity  when  a  definite  ratio  of  intensity  between  a  given 
light-source  and  the  surrounding  field  is  maintained.  I  hope 
that  later  both  of  these  points  will  be  investigated.  On  the  latter 
point  I  have  at  present  this  much  evidence  to  offer.  When 
working  with  either  a  direct  or  semi-indirect  system  of  lighting, 
the  eye  both  falls  off  less  in  efficiency  and  experience  less  dis- 
comfort at  the  lower  scales  of  intensity  than  at  the  higher.  The 
most  favorable  intensity  too  is  found  to  be  less  than  that  at- 
tained by  Dr.  Ives  with  his  artificial  window.  If  it  can  be 
assumed,  then,  that  the  ratio  of  intensity  of  source  to  surround- 
ing field  in  our  work  remained  approximately  constant  as  the 
general  scale  of  intensity  was  changed,2  as  I  have  every  reason 
to  believe  it  did,  the  above  results  would  lead,  so  far  as  the  case 
is  representative,  to  a  conclusion  which  is  just  the  reverse  of 
that  suggested  by  Dr.  Ives  in  partial  explanation  of  the  excess 
of  discomfort  caused  by  the  artificial  window.  I  would  in  fact 
myself  be  very  much  inclined  to  seek  further  for  a  factor  in 
the  cause  of  the  discomfort.  Dr.  Ives  secured  his  distribution 
by  a  number  of  reflecting  plates  set  at  different  angles  with 
small  spaces  between.  The  reflection  from  these  plates  was, 
moreover,  only  partially  diffuse.  To  the  eye  in  any  given  posi- 
tion, the  surface  brilliancy  of  the  plates  was  not  by  any  means 
uniform  nor  even  uniformly  graded  from  point  to  point.     Quite 

2  Ifi  these  experiments  clear  tungsten  lamps  ranging  in  wattage  from  15  to  100  were 
used.  In  each  test  the  wattage  from  fixture  to  fixture  was  uniform,  ;.  e.,  the  lamps  in 
all  the  fixtures  were  either  15's,  25's,  40's,  6o's,  or  100's.  depending  on  the  intensity  de- 
sired. 


258  TRANSACTIONS    I.  E.  S. — PART    II 

considerable  specular  reflection  and  glare  were  present.  Viewed 
by  the  eye  in  any  given  position,  the  window  had  more  or  less 
the  appearance  of  several  light-sources  of  different  intensities. 
The  surface  brilliancy  of  a  window  illuminated  by  daylight  seems 
on  the  other  hand,  to  the  unaided  eye  at  least,  to  be  more  uni- 
form, or  at  least  more  uniformly  graded.  The  surfaces  that 
reflect  the  light  into  the  room,  corresponding  to  Dr.  Ives's  plates, 
— the  sky,  the  pavement,  the  walls  of  neighboring  buildings,  etc., 
are  in  general  not  in  the  field  of  vision  or  at  least  not  so  largely 
in  the  field  as  are  Dr.  Ives's  plates.  Moreover,  the  reflecting 
surfaces  which  are  in  general  concerned  in  the  illumination  of  a 
room  by  daylight  give  diffuse  reflection  and  not  specular.  In 
short,  it  seems  that  something  that  affects  the  comfort  of  the  eye 
has  escaped  Dr.  Ives  in  his  photometric  analysis  of  distribution. 
Whether  or  not  it  is  something  that  could  be  detected  by  photo- 
metric analysis  I  am  not  prepared  to  say.  It  would  seem  to  me, 
however,  that  if  an  adequate  check  were  to  be  had  in  both 
cases  on  distribution  of  light  in  the  room  and  of  surface  bril- 
liancy in  the  source,  readings  would  have  had  to  be  taken  in  a 
greater  number  of  directions  than  were  taken  by  Dr.  Ives.  That 
is,  it  seems  to  me  possible  to  have  duplicated  by  means  of  re- 
flecting plates,  especially  plates  which  were  not  completely  dif- 
fusely reflecting,  set  at  appropriate  angles,  the  readings  in  the 
number  of  directions  he  employed  and  still  not  have  had  a  dis- 
tribution and  surface  brilliancy  by  any  means  identical  with  that 
present  in  his  illumination  by  means  of  the  daylight  window. 

Dr.  H.  E.  Ives  (communicated)  :  Upon  seeing  the  amplified 
discussion  which  Dr.  Ferree  has  submitted  in  writing,  I  feel  it 
encumbent  on  me  to  describe  more  in  detail  some  of  the  experi- 
mental conditions  and  correct  what  I  believe  are  misconceptions 
on  his  part.  In  general  Dr.  Ferree  believes  that  the  third  copy 
of  daylight  described  is  not  as  complete  as  the  paper  would  lead 
one  to  believe, — that  differences  are  present  that  escape  the 
photometer.  I  think  that  Dr.  Ferree  has  not  properly  grasped 
the  fact  that  the  measurements  given  are  not  of  illumination. 
but  of  surface  brightness.  Thus,  when  he  speaks  of  the  "most 
favorable  intensity,"  as  determined  by  his  own  experiments,  being 
less  than  that  of  these  experiments,  he  can  be  giving  only  an  esti- 


SOME   HOME  EXPERIMENTS  IN   ILLUMINATION  259 

mate  of  intensity  of  illumination,  since  the  intrinsic  brightness 
of  the  direct  and  semi-indirect  units  being  studied  by  him  is  tens 
or  hundreds  of  times  greater  than  the  brightest  visible  object  in 
the  window  experiment.  In  fact,  so  extremely  different  in  mag- 
nitude and  distribution  are  the  visible  bright  areas  in  Dr.  Ferree's 
experiments  and  mine  that  it  appears  to  me  out  of  the  question 
to  attempt  any  comparison  based  on  a  mere  guess  at  the  relative 
illuminations.  Considerable  stress  is  laid  by  him  on  the  "specular 
reflection"  of  the  plates  constituting  the  artificial  windows.  My 
description  of  these  as  partly  specular  was  unfortunate,  as  it 
gives  the  idea  of  reflected  images.  A  better  description  would  be 
"matt  reflecting  surfaces  of  varying  reflection  coefficient  in  differ- 
ent directions."  The  sandblasted  mirrors  as  used  give  reflected 
"images"  of  perhaps  one  foot  diameter.  The  light  sources  were 
six  inches  apart;  consequently  it  is  easily  possible  to  arrange 
these  reflectors  to  present  a  surface  entirely  uniform  in  bright- 
ness. It  is  physically  impossible,  when  care  is  taken  to  secure 
perfect  uniformity  of  reflecting  power  in  the  various  elements, 
to  distinguish  by  inspection  from  any  given  direction  that  the 
reflection  is  not  perfectly  matt. 

It  is  quite  true  that  the  different  sheets  were  not  uniformly 
bright,  but  this  was  because  before  Dr.  Ferree  saw  the  installation 
absolute  uniformity  was  tried  out  and  found  far  more  trying  to 
the  eye  than  a  certain  amount  of  irregularity,  which  was.  there- 
fore, allowed  to  creep  in.  This  latter  is  decidedly  less  than  that 
present  by  day  in  the  landscape  seen  through  the  window. 

I.  therefore,  hold  to  my  claim  that  the  conditions  produced 
were  a  very  accurate  copy  of  the  daylight  conditions  and  appear- 
ance, with  the  exceptions  noted,  namely,  the  proximity  of  the 
bright  surfaces,  the  low  absolute  intensity  and  the  color  of  the 
light.  I  entirely  agree  with  Dr.  Ferree  tkat  these  factors  ought 
to  be  tested  out  separately  to  determine  their  relative  importance. 
We  look  to  the  psychologist  to  do  this. 

Mr.  Robert  B.  Ely  :  Decorators  have,  in  numerous  instances, 
used  cretonne  screens  on  candles  in  dining  rooms  and  bed  rooms, 
etc.,  so  constructed  as  to  reflect  the  light  on  the  wall,  the  shades 
being  elliptical  in  shape  and  so  placed  as  to  conceal  the  light 


260  TRANSACTIONS    I.  E.  S. — PART    II 

source  from  view.  In  a  great  many  of  these  instances  I  have 
found  that  there  has  been  complaint  of  poor  illumination  in  the 
center  of  the  room  with  the  candles  equipped  in  this  manner. 
In  bed  rooms  this  has  been  overcome  by  placing  two  small  port- 
able lamps  equipped  with  shades  on  bureaus  or  chiffoniers,  so 
that  the  light  was  reflected  from  bureau  scarf  to  illuminate  the 
face  of  the  person  standing  in  front  of  the  bureau  or  chiffonier. 
In  dining  rooms  it  has  been  necessary  to  resort  to  the  use  of  a 
candelabra,  equipped  with  imitation  candles  and  suitable  shades. 

Mr.  R.  L.  Lloyd:  It  seems  to  me  the  natural  direction  for 
light  is  from  overhead.  We  are  living  under  artificial  conditions 
in  houses  with  windows  in  the  side  walls,  and  have  become  ac- 
customed to  seeing  light  enter  that  way,  but  the  natural  tendency 
will  generally  be  found  to  be  to  turn  one's  back  to  the  light,  so 
as  to  keep  it  out  of  the  eyes.  When  the  Ancients  first  began  to 
utilize  artificial  light  they  turned  naturally  toward  locating  it 
above,  and  I  understand  that  some  of  their  temples  were  made 
without  roofs,  so  that  light  could  enter  in  the  natural  direction. 

Although  these  experiments  of  Dr.  Ives  are  valuable  as  re- 
searches in  science,  I  think  he  is  working  in  the  wrong  direction, 
in  trying  so  assiduously  to  imitate  lighting  as  entering  from 
windows,  when  it  would  be  much  more  easily  accomplished  to 
arrange  the  light  to  come  from  above.  We  have  seen  by  the 
demonstrations  of  Mr.  Luckiesh  and  others  that  when  the  light 
is  directed  on  natural  objects  from  above,  we  see  them  in  their 
natural  appearance,  and  that  when  the  light  is  directed  from 
other  than  above,  their  appearance  is  distorted.  The  photo- 
grapher too  has  learned  this,  and  always  makes  use  of  a  skylight 
in  his  studio  for  taking  pictures. 

As  stated  before,  these  experiments  of  Dr.  Ives  are  very 
interesting  and  are  valuable  contributions  to  our  science,  but  I 
have  yet  to  be  convinced  that  a  simulation  of  artificial  conditions 
is  a  proper  one  for  the  best  results. 

Prop.  A.  J.  Rowland:  I  have  had  the  privilege  of  seeing  the 
installation  of  lights  and  lighting  referred  to  in  Dr.  Ives's  paper. 
I  must  confess  to  a  certain  first  sense  of  bewilderment  when  I 
saw  Dr.  Ives's  installation,  especially  that   for  producing  day- 


SOME   HOME  EXPERIMENTS  IN   ILLUMINATION  26l 

light  values.  I  am  so  used  to  seeing  lights  in  certain  places, 
mounted  in  a  conventional  way,  that  I  have  pretty  strong  feel- 
ings about  something  special  or  unusual.  The  simplicity  of  the 
side  wall  lighting  scheme,  and  the  good  results  secured  from  it, 
impressed  me  strongly.  When  it  can  be  used  intelligently,  it 
seems  to  me  to  be  one  with  points  of  great  merit.  The  trouble 
with  most  plans  which  secure  good  lighting  by  the  use  of  things 
outside  the  lighting  fixture  itself  as  a  source,  comes  from  the 
extraordinary  way  in  which  applications  are  made  by  light  users. 
Dr.  Ives's  walls  are  covered  in  such  a  way  that  they  lend  them- 
selves splendidly  to  use  as  secondary  light  sources.  At  my  home 
a  deep  green  felt  paper  would  make  the  plan  worthless.  I 
wonder  whether  in  ordinary  service  most  people  would  not 
swing  the  shade  180  degrees  from  its  proper  position  and  utterly 
spoil  the  lighting  scheme.  I  like  the  large  source  plan;  I  like 
the  idea  of  placing  the  main  sources  of  light  on  levels  similar 
to  those  from  which  daylight  is  derived;  and  the  simplicity  of 
the  whole  side  lighting  plan  has  much  to  commend  it. 

In  connection  with  the  paper,  the  term  "brightness"  has  been 
used  a  number  of  times.  I  think  it  is  one  which  needs  to  be 
explained  carefully.  If  one  compares  the  brightness  of  a  white 
wall  a  gray  wall,  or  a  dark  colored  floor,  it  is  a  thing  independent 
of  color.    Just  what  does  it  imply?    How  is  it  to  be  measured? 

Mr.  Charles  O.  Bond:  The  semi-indirect  side-wall  lighting 
scheme  proposed  and  favored  by  Dr.  Ives  deserves  particular 
attention.  He  shows  that  strong  contrasts  in  the  field  of  vision 
are  avoided ;  that  the  method  lends  itself  to  artistic  treatment  and 
adornment  of  a  room  into  which  even  color  variations  may  be 
introduced ;  and,  best  of  all,  the  method  has  "worn  well." 

His  data  have  been  obtained  in  a  home,  and  in  such  conditions 
fixed  positions  are  not  compulsory.  If  the  lighting  is  found  try- 
ing while  one  is  occupying  one  seat,  it  is  easy  to  change  to 
another.  That  is  one  reason  why  such  indefensible  lighting  pre- 
vails in  many  houses. 

The  side-wall  semi-indirect  method  ought  to  be  easily  adaptable 
with  greatly  increased  resultant  comfort  to  small  audience  cham- 
bers, such  as  church  parlors,  where  persons  may  not  easily  change 
the  lighting  conditions  through  a  change  of  seat. 


262  TRANSACTIONS    I.  E.  S. — PART    II 

Dr.  H.  E.  Ives  (In  reply)  :  I  have  nothing  to  add  to  the 
discussion.  My  hope  is  that  this  paper  may  direct  thought  to 
certain  other  possibilities  in  lighting  than  are  now— due  to  effi- 
ciency considerations — most  common. 


PIERCE:     GASLIGHTINC    IN    AN     EXHIBITION    HALL  263 

GASLIGHTING   IN  AN  EXHIBITION  HALL.* 

HV    ROBERT    F.    PIERCE. 


Synopsis:  The  following  article  describes  a  temporary  semi-indirect 
gas  lighting  installation  which  was  provided  for  an  exhibition  hall.  Ten 
lighting  units,  each  consisting  of  fifteen  upright  burners  within  an  orna- 
mental glass  bowl  mounted  on  a  pedestal,  furnished  the  illumination.  The 
illustrations  show  the  plan  of  the  unit,  a  diagram  and  a  night  view  of 
the  interior. 

The  illumination  of  Taft  Hall  in  the  Auditorium  Armory, 
Atlanta,  Ga.,  is  the  result  of  a  number  of  compromises  with 
more  or  less  unfavorable  elements,  rather  than  the  unrestrained 
working  out  of  a  consistent  and  coherent  plan,  and  must  be 
judged  in  the  light  of  the  existing  circumstances.  The  lighting 
was  primarily  designed  for  use  in  connection  with  the  conven- 
tion of  the  National  Commercial  Gas  Association.  The  same 
room  was  to  be  used  for  the  beefsteak  dinner  with  which  the 
convention  was  terminated,  and,  in  addition  to  this,  it  was  desired 
to  furnish  a  lighting  system  which  should  be  available  for  the 
automobile  show  which  preceded  the  gas  convention. 

The  problem  then  resolved  itself  into  meeting,  as  well  as  pos- 
sible, the  requirements  imposed  by  each  of  these  widely  varying 
purposes  with  one  installation.  Naturally,  it  was  desired  that  the 
installation  embody  unique  features— at  least,  features  unique  as 
far  as  gas  lighting  is  concerned — and  it  was  suggested  that  indi- 
rect lighting  be  employed  in  connection  with  pedestals ;  something 
along  the  line  of  the  installation  in  the  Louis  XVI  dining  room 
in  the  Congress  Hotel,  Chicago. 

It  so  happened  that  the  room  was  peculiarly  adapted  to  the 
support  of  the  lighting  units  on  columns  or  pedestals.  In  fact, 
this  was  the  only  feasible  arrangement.  The  ceiling  is  plain  and 
unbroken,  except  by  four  beams  which  divide  off  a  large  square 
in  the  center  of  the  ceiling  comprising  nearly  one-half  the  ceiling 
area.  This  square  is  further  sub-divided  by  a  single  longitudinal 
beam  of  the  same  dimensions  as  the  others.  Obviously  the  ceil- 
ing presented  no  location  for  the  suspension  of  fixtures. 

The  side  walls  were  devoid  of  pilasters  or  similar  features  to 

*  A  paper  read  before  a  meeting  of  the  New  York  section  of  the  Illuminating:  Engi- 
neering Society,  January  9,  1913. 

5 


264  TRANSACTIONS   I.  E.  S. — PART   II 

which  wall  brackets  might  have  been  attached,  and  the  only 
remaining  expedient  was  in  the  use  of  columns  for  the  support 
of  the  lighting  units. 

For  the  lighting  of  the  hall  for  convention  purposes,  the  best 
arrangement  of  the  columns  appeared  to  be  as  shown  in  Fig.  5, 
marking  off  an  area  bounded  by  the  supporting  columns  and 
fronting  upon  the  rostrum  from  which  the  convention  was  con- 
ducted. This  arrangement  also  served  very  well  for  the  lighting 
of  the  automobile  show,  the  booths  being  arranged  in  such  a 
manner  that  the  lighting  columns  marked  the  rear  corners  of 
the  booths  when  viewed  from  the  middle  aisle,  and  the  front 
corners  when  viewed  from  the  side  aisles. 

In  the  design  of  the  lighting  columns,  it  was  considered  desir- 
able to  depart  from  the  Ionic  order  of  the  supporting  columns  for 
two  reasons.  In  the  first  place,  the  use  of  similar  columns  would 
have  given  them  the  appearance  of  being  original  structural 
members,  decapitated  for  the  purpose  of  bearing  the  lighting 
units,  whereas,  it  was  felt,  a  different  treatment  would  separate 
the  lighting  columns  into  a  distinct  system,  having  its  own  reason 
for  existence.  In  the  second  place,  it  was  desired  to  utilize  a 
design  which  would  not  be  out  of  place  in  other  surroundings, 
as  it  was  likely  that  the  installation  would  be  sold  for  other  pur- 
poses after  the  convention.  The  composite  Corinthian  order  was 
selected  as  having  the  widest  field  of  possible  future  applications 
and  one  which  would  harmonize  fairly  well  with  any  classic 
interior. 

In  the  design  of  the  light  distribution  the  direct,  semi-indirect 
and  indirect  systems  received  consideration.  The  latter,  origin- 
ally suggested,  had  a  number  of  serious  drawbacks.  Any  system 
of  illumination  which  turns  the  room  optically  upside  down  by 
reason  of  reversing  the  natural  order  of  intensities  increasing 
from  the  ceiling  downward,  is  certain  to  be  a  source  of  more  or 
less  definite  annoyance,  especially  where  the  attention  of  the 
occupants  of  the  room  is  not  apt  to  be  concentrated  upon  work 
which  distracts  their  attention  from  the  abnormal  distribution  of 
illumination  in  the  room.  This  objection  might  be  of  little 
moment  in  a  room  devoted  to  clerical  purposes,  but  in  the  hall 
under  consideration,  was  held  to  be  of  much  importance.     The 


*/> 


Fig.  i. — View  of  interior  of  hall. 


Fig.  2.— Design  of  bowl 
and  pedestal. 


Fig.  3. — Plan  of  light  unit,  showing 
arrangement  of  lamps. 


Fig.  4. — Distribution  curve  from  light  unit.  In  the  above  distribution  curve 
the  radii  vectores  of  the  right-hand  curve  represents  the  fluxes  in  zones  ex- 
tending 5  degrees  on  either  side  of  the  designated  angle,  while  those  of  the 
right-hand  curve  represent  the  total  flux  from  zero  to  the  designated  angles. 


Fig.  5. — Plan  of  interior  of  the  hall. 


PIERCE:     GASUGHTING    IN    AN    EXHIBITION    HALE  265 

absence  of  an  apparent  source  of  light  was  deemed  a  considerable 
drawback  to  the  indirect  system  from  a  similar  consideration. 
Furthermore,  only  the  presence  of  a  luminous  light  source  could, 
from  an  esthetic  standpoint,  provide  any  excuse  for  the  existence 
of  the  lighting  columns. 

For  the  lighting  of  the  automobile  show  it  was  felt  that  the 
presence  of  luminous  light  sources  was  especially  desirable,  as 
the  appearance  of  an  abundance  of  light  seems  to  be  an  indis- 
pensable feature  of  exhibitions  of  this  kind. 

The  foregoing  considerations  seemed  to  narrow  the  choice  to 
the  selection  of  either  direct  or  semi-indirect  lighting. 

For  the  automobile  show,  an  illumination  of  about  8  lumens 
per  square  foot  was  deemed  desirable  on  account  of  the  light- 
absorbing  qualities  of  the  black  car-bodies.  The  area  of  the 
hall  (about  5,900  square  feet)  demanded,  therefore,  a  horizontal 
component  of  about  47,000  lumens,  which  demanded,  at  a  utiliza- 
tion efficiency  of,  say  40  per  cent,  (the  walls  are  dark  green  in 
color)  about  120,000  lumens  generated.  It  seemed  desirable  to 
limit  the  number  of  lighting  units  to  ten  in  order  that  they  might 
not  interfere  with  the  best  use  of  the  floor  space.  This  demanded 
the  production  of  12,000  lumens  per  unit.  One  candle-power  per 
square  inch  was  decided  upon  as  an  upper  limit  for  the  specific 
intensity  of  the  bowls  or  globes.  This  would  have  required  a 
globe  of  a  size  altogether  out  of  proportion  with  the  columns, 
and  it  was  finally  decided  to  use  semi-indirect  lighting. 

An  indirect  component  of  50  per  cent,  was  decided  upon  as 
meeting  the  various  requirements  to  the  most  satisfactory  degree. 
This  ensured  the  reduction  of  shadow  contrasts  to  a  point  well 
within  the  requirements  of  the  automobile  show,  where  ample 
diffusion  of  light  for  the  inspection  of  machine  interiors  was 
required  and  at  the  same  time  avoided  the  "flatness"  which  would 
have  accompanied  a  much  lower  indirect  component.  It  was 
believed  that  the  50  per  cent,  direct  component  would  be  sufficient 
to  bring  the  apparent  plane  of  highest .  illumination  well  down 
toward  the  floor,  where  it  belonged. 

A  light  distribution  (Fig.  4)  was  finally  decided  upon  having 
an  extensive  character  above  the  horizontal,  with  a  maximum 
at  135  degrees.     With  the  units  as  located,  this  provided  sub- 


266  TRANSACTIONS    I.  E.  S. — PART    II 

stantially  uniform  ceiling  illumination,  preserving  the  flatness  of 
the  ceiling.  The  ceiling  being  a  very  light  cream  in  color,  it  was 
assumed  that  about  50  per  cent,  of  the  light  from  no  degrees 
upward  would  be  effective  on  the  working  plane,  giving  about 
30,000  effective  lumens  from  the  ceiling.  Assuming  70  per  cent, 
as  the  effective  angle  below  the  horizontal,  30,000  lumens  would 
be  contributed  to  the  working  plane  direct  from  the  lamps,  giving 
an  average  effective  illumination  of  about  ten  lumens  per  square 
foot — about  50  per  cent,  of  which  is  indirect. 

The  bowls  were  primarily  designed  to  furnish  this  curve  by 
reflection  from  the  glass  itself,  and  to  harmonize  the  form  of 
the  bowl  with  the  column,  it  was  treated  as  a  conventionalized 
floral  form,  borne  by  the  conventionalized  tree  trunk  and  foliage 
which  comprised  the  shaft  and  capitol  of  the  columns.  For  this 
reason,  it  was  necessary  to  depart  from  the  strictly  classic  orders 
in  designing  the  bowl.  Equalite  glass  was  selected  for  the  bowl 
on  account  of  its  shell-like  texture  and  selective  absorption,  giving 
a  warm  lively  glow  in  the  more  transparent  of  the  irregular  stria- 
tions  of  its  structure.  This  relieved  the  cold,  severe  white  of  the 
denser  portions  without  conflicting  with  the  material  of  the  col- 
umns, which  was  a  concrete  composed  of  cement  and  marble  dust. 
The  necessity  of  avoiding  colors  in  the  bowl  itself  imposed  the 
use  of  conventionalized  floral  forms  for  decoration  which  should 
be  devoid  of  any  suggestion  of  color,  and  the  lotus  was  selected 
for  this  purpose  as  harmonizing  both  in  form  and  color  with  the 
bowl.  This  might  be  criticised  as  an  injudicious  admixture  of 
Greek  and  Egyptian  orders,  but  it  was  felt  that  such  criticism 
is  captious,  there  being  no  necessity  for  adhering  to  purely 
geographical  distinctions  in  this  instance.  The  design  of  bowl 
and  column  finally  selected  is  shown  in  Fig.  2. 

Fig.  1  was  made  from  a  photograph  of  the  completed  installa- 
tion and  conveys  a  fair  idea  of  the  distribution  of  illumination. 
It  appears  from  this  picture  that  the  indirect  illumination  is 
rather  overdone.  The  dark  floor,  however,  contributed  a  great 
deal  toward  this  impression. 

It  was  exceedingly  unfortunate  that  the  ceiling  fixtures  and 
draperies  could  not  have  been  removed.  They  were  extremely 
offensive  and  marred  what  would  otherwise  have  been  a  quite 


PIERCE:     GASLIGHTING    IN    AN    EXHIBITION    HALL  267 

agreeable  interior.  It  was  also  unfortunate  that  the  interior  could 
not  be  redecorated  in  such  a  way  as  partly  to  correct  for  the  inver- 
sion of  illumination  intensities  by  a  compensating  color  gradation. 
Had  this  been  possible,  however,  it  is  doubtful  if  it  would  have 
availed  much  with  the  high  intensities  made  necessary  for  the 
lighting  of  the  automobile  show. 

Fig.  3  shows  the  arrangement  of  lamps  in  the  bowls.  Fifteen 
upright  burners  are  mounted  in  two  concentric  rings  and  supplied 
with  gas  through  magnet  valves.  Ignition  is  accomplished  by  a 
flash  pilot,  throwing  a  jet  under  an  inverted  annual  trough  above 
the  burner  chimneys.  This  pilot  is  simply  an  automatic  open- 
flame  lighter,  or  Boston  cock,  which  is  lighted  by  a  make-and- 
break  spark — an  arrangement  familiar  to  everyone  who  has  used 
open-flame  gas  burners. 

The  entire  burner  arrangement  is  composed  of  standard  parts 
and  appliances  which  have  been  used  for  many  years  with  entirely 
satisfactory  results,  so  that  this  installation  is  in  no  sense  a  special 
one  without  commercial  utility,  but  exemplifies  possibilities  in  the 
way  of  securing  unique  and  pleasing  effects  which  may  be  utilized 
by  any  designer  of  gas  lighting  installations  who  cares  to  avail 
himself  of  them. 


268  TRANSACTIONS   I.  %.  S. — PART   II 

METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING.* 


THOMAS    W.   ROIvPH. 


Synopsis:  After  brief  introductory  comments  on  the  progress  in 
industrial  lighting,  the  author  of  this  paper  presents  a  classification  of 
metal  reflectors  and  a  discussion  of  the  nature  of  reflection.  For  the 
most  part,  the  paper  is  devoted  to  the  merits  and  demerits  of  porcelain- 
enamel  and  aluminum-finished  steel  reflectors  and  a  consideration  of  the 
distribution  characteristics  which  influence  the  selection  of  the  different 
types  of  these  reflectors  for  industrial  lighting  installations.  The  cost 
of  reflectors,  the  author  states,  is  of  minor  importance  in  that  it  is 
readily  compensated  for  by  effective  illumination  when  the  lighting  system 
is  properly  designed. 

Up  to  the  present  time,  the  progress  of  industrial  lighting  has 
followed  the  progress  of  lighting  in  the  general  field.  In  effi- 
ciency, in  variety  and  excellence  of  distribution,  in  diffusion  and 
in  eye-protection,  the  improvement  of  reflectors  for  use  in  the 
general  lighting  field  has  preceded  the  improvement  of  industrial 
lighting  reflectors.  We  have  now  reached  a  point  at  which  this 
condition  is  likely  to  change.  Industrial  lighting  is  assuming  a 
greater  importance  and  it  is  not  at  all  improbable  that  some  of 
the  most  noteworthy  advances  in  illuminating  engineering  in  the 
next  few  years  will  originate  in  this  field.  The  reason  is  not 
hard  to  find.  This  is  a  commercial  age  and  whatever  can  be 
reduced  to  dollars  and  cents  will  receive  the  maximum  amount 
of  attention  from  the  business  men  of  the  country;  wherever  im- 
provements can  be  shown  to  affect  profits  most  vitally,  there  im- 
provements will  be  most  rapid.  In  the  lighting  of  factories,  good 
lighting — meaning  adequate  intensity  of  illumination,  proper 
protection  of  the  eyes  and  a  high  illumination  efficiency — can  be' 
shown  to  affect  profits  more  directly  and  to  a  greater  degree  than 
in  the  general  field  of  commercial  lighting. 

Those  interested  in  the  advancement  of  industrial  lighting 
have  before  them  the  task  of  disseminating  among  a  large  body 
of  variously  employed  individuals  the  underlying  principles  of 
good  illumination.  Factory  engineers  and  managers,  who  as  a 
class  have  already  shown  their  appreciation  of  good  lighting, 

*  A  paper  read  at  a  meeting  of  the  Philadelphia  section  of  the  Illuminating  fingineer- 
ing  Soeiety,  May  16,  1913. 


R0LPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      269 


salesmen  who  sell  all  kinds  of  reflectors,  bad  as  well  as  good,  gas 
and  electric  companies'  solicitors,  who  are  expected  to  sell  results 
in  illumination  as  well  as  gas  or  electricity ;  all  these  should  have 
a  clear  understanding  of  what  constitutes  good  illumination  and 
of  how  reflectors  control  light  in  producing  it.  The  proper  redi- 
rection of  the  light  of  the  lamp  by  means  of  reflectors  is  essen- 
tial to  good  lighting.  A  knowledge  of  the  merits  and  demerits  of 
the  reflectors  available  for  factory  lighting  is,  therefore,  of  par- 
ticular importance. 

Metal  reflectors  for  industrial  lighting  may  be  classified  as 
follows : 


f   Material 


Character  of 
Reflection 


Steel 
Brass 
Aluminum 

(   Specular 
Spread 


Metal 
Reflectors 


L 


Diffuse 


•(    Polished  metals 

f   Rough  metal 
surfaces 
Applied 
aluminum 


Porcelain 
enamel 
Paint  enamel 


Shape 


Distribution  of 
Light 


f    Deep  bowl 
j    Shallow  bowl 
]    Shallow 
^   Angle 


Extensive 
Intensive 
■{    Focusing 
I    Distributing 
L   Asymmetric 


The  metal  most  widely  used,  and  rightly  so,  is  steel.  It  is 
durable,  reasonably  low  in  price  and  readily  takes  an  applied 
finish  of  almost  any  character.  Brass  and  aluminum  are  both 
more  expensive  than  steel  and  somewhat  less  durable.  Brass 
finds  a  limited  field  of  usefulness  in  reflectors  of  such  shapes  that 
cannot  be  drawn  or  spun  of  steel.  Aluminium  has  a  slight  ad- 
vantage because  it  is  light  in  weight  and  its  surface,  when  not 
polished,  is  excellent  from  the  standpoints  of  efficiency  of  re- 
flection and  character  of  illumination  produced. 

The  character  of  reflection  obtained  from  a  metal  reflector 
is  very  important  in  determining  the  value  of  a  reflector.     By 


270  TRANSACTIONS    I.  E.  S. — PART    II 

character  of  reflection  is  meant  the  character  of  the  action  of  the 
surface  upon  each  minute  pencil  of  light-rays,  i.  e.,  whether  it  re- 
flects the  pencil  regularly  changing  its  direction  only,  or  whether 
it  breaks  up  the  pencil,  reflecting  light  in  many  different  direc- 
tions. There  is  a  very  general  lack  of  understanding  of  these 
actions.  The  law  that  the  angle  of  reflection  is  equal  to  the  angle 
of  incidence,  i.  e.,  that  a  light-ray  is  always  reflected  at  the  same 
angle  with  the  surface  as  the  angle  at  which  it  strikes  (Fig.  i-a), 
is  given  credit  for  a  much  wider  field  of  usefulness  than  it  really 
has.  This  law  holds  for  every  reflection  of  an  individual  light- 
ray  by  an  infinitesimal  portion  of  a  surface ;  but  many  surfaces 
are  rough  and  many  other  surfaces  allow  light  to  pass  into  them 
and  be  reflected  from  particles  beneath,  so  that  light-rays  which 
are  parallel  when  striking  a  surface,  are  often  broken  up  and 
reflected  in  many  directions. 

Reflecting  surfaces  in  common  use  may  be  grouped  into  three 
distinct  classes,  according  to  the  manner  in  which  they  reflect 
light.  These  three  varieties  of  reflection  are  specular  reflection, 
spread  reflection  and  diffuse  reflection.  Fig.  1  shows  the  char- 
acter of  each  of'these.  In  each  case,  the  dotted  line  surrounding 
the  reflected  rays,  is  what  might  be  termed  the  photometric  curve 
of  the  light  reflected  from  any  point  on  the  surface.  Specular 
reflection  rigidly  follows  the  law  that  the  angle  of  reflection  is 
equal  to  the  angle  of  incidence.  The  reflecting  surface  is  smooth 
and  the  reflected  ray  always  makes  the  same  angle  with  the  sur- 
face as  the  incident  ray.  In  spread  reflection  the  maximum 
candle-power  of  reflected  light  is  in  the  same  direction  as  in 
specular  reflection.  The  light  is  broken  up,  however,  and  slightly 
spread  from  the  direct  path.  In  diffuse  reflection  the  angle  of' 
incidence  has  no  effect  upon  the  reflected  light.  No  matter  at 
what  angle  the  light  strikes  the  surface,  the  maximum  reflected 
ray  is  normal  to  the  surface  and  the  light  is  reflected  in  all 
directions  in  accordance  with  the  well-known  cosine  law.  The 
photometric  curve  of  each  point  on  the  surface  is  a  tangent 
circle.  Fig.  2  illustrates  the  manner  in  which  each  of  these  three 
kinds  of  reflection  is  produced  by  the  infinitesimal  portions  of  the 
surfaces.  Regular  reflection  is  produced  by  any  smooth  opaque 
surface.     Mirrors  and  polished  metals  produce  regular  reflection. 


\r 


tp,* 


Fig.  i.— Character  of  reflection  obtained  from  various  surfaces, 
(a)— Specular  reflection;  (b)—  spread  reflection;   (c)— diffuse  reflection. 


o  d  °   °    - 
Fig.  2. — Reflecting  action  of  various  surfaces. 
a       Regular  reflection;  (b) — irregular  reflection;  (c) — sub-surface  reflection. 


Fig.  ?. — Combination  of  diffuse  and  specular  reflection. 


s 

S  SSS3 

s 

s     s 

1  (  1  1  1  lnr~~j^M 

JpX~^\  \  \-\-X\ 

'          Ul 

Xt  _-'--•!  o- 

c 

LJ/  J  1  1 

>CJ~8i0"" 

)*OpriZ0- 

v^T"°-i60- 

ciomjjAlone     X/ 
-Lompw.fhR.rflecJpr^ 

Fig.  4. — Extensive  aluminum  finished 
steel  reflector.  (Photometric  curve 
shown  in  Fig.  5.) 


Fig.  5. — Photometric  curve  of  extensive 
aluminum  finished  steel  reflector,  with 
100-watt  clear  tungsten  lamp  operating 
at  1. 13  watts  per  candle.  (Reflector  is 
shown  in  Fig.  4.) 


Fig.  6.— Shallow  type  distributing  steel 
reflector,  porcelain  enameled.  (Photo- 
metric curve  shown  in  Fig.  7.) 


Fig.  8. — I,arge  size  angle  reflector,  porce- 
lain enameled  steel.  (Photometric 
curve  shown  in  Fig.  9.) 


j — l  y^v\>^  \ 

''1  TiujIl 

/    '. 

1    1      M 

1  v^*-"***!^  '■-■*•'•'    . 

xf\y~cr/s"  1^0 — y 

Vv^Cj^SkT" 

Vv 

^\      ?• — 300 

ti«y^/ 

Fig.  7. — Photometric  curve  of  shallow 
type  of  porcelain  enameled  steel  re- 
flector with  150-watt  clear  tungsten 
lamp  operated  at  1.13  watts  percandle. 
(Reflector  shown  in  Fig.  6.) 


'  v  V'V"- 

d$ 

AXV 

VLar 

np  Alor.e\^\.-- 
injMthRefledor 

V^f~7 

x>3v 40°  \ 

///f*^l 

^  75—800  - 

^^T 1000  - 

<x// 

150° 


Fig.  9.— Photometric  curve  of  large  size 
porcelain  enameled  angle  steel  reflec- 
tor with  500-watt  clear  tungsten  lamp 
operating  at  1.00  watt  per  candle.  (Re- 
flector shown  in  Fig.  8.) 


ROLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      2J  I 

Irregular  reflection  of  the  individual  light-rays  (Fig.  2-b)  causes 
a  spreading  of  the  light,  resulting  in  spread  reflection.  This  is 
produced  by  any  matte  surface.  The  surface  of  this  character 
most  widely  used  for  reflectors  is  the  applied  aluminum  finish. 
Diffuse  reflection  is  almost  invariably  produced  by  reflection  of 
light-rays  from  particles  beneath  the  surface  (Fig.  2-c).  The 
most  widely  used  surface  of  this  character  is  porcelain  enamel. 
Porcelain  enamel  is,  in  reality,  a  glass  having  the  characteristics 
of  opal.  The  minute  particles  held  in  suspension  in  the  glass 
are  the  reflecting  media.  In  addition  to  the  light  reflected  from 
these  particles  beneath  the  surface,  there  is  a  small  amount  of 
reflection  from  the  surface  itself.  This  reflection  is  regular  since 
the  surface  is  smooth.  The  result  is  a  combination  of  diffuse 
and  specular  reflection  as  shown  in  Fig.  3.  The  light  reflected 
specularly  is  usually  negligible  in  quantity,  but  cannot  be  entirely 
overlooked,  especially  in  reflector  design. 

Surfaces  giving  specular  reflection  afford  a  high  degree  of 
light  control  and  polished  metal  reflectors  could  be  designed  to 
give  almost  any  desired  distribution  of  light.  Such  surfaces  are 
very  little  used,  however,  due  to  the  streaked  character  of  the 
illumination  obtained  from  them.  The  polished  aluminum  re- 
flector, used  considerably  several  years  ago.  is  now  well-recog- 
nized as  a  highly  undesirable  reflector  because  of  the  striations 
or  streaks  in  the  resulting  illumination.  Surfaces  giving  spread 
reflection  are  only  slightly  less  susceptible  to  accurate  design  than 
surfaces  giving  specular  reflection  since  with  spread  reflection, 
the  major  portion  of  the  reflected  light  deviates  only  slightly 
from  the  law  that  the  angle  of  reflection  is  equal  to  the  angle  of 
incidence.  Aluminum  finished  reflectors,  for  example,  can  be 
accurately  designed  to  give  all  the  most  widely  useful  kinds  of 
light-distribution.  Diffusely  reflecting  media,  however,  such  as 
porcelain  enamel  and  paint  enamel,  have  a  much  narrower  field 
of  usefulness.  Extreme  distributions  pf  light,  such  as  the  focus- 
ing type,  cannot  be  obtained  with  reflectors  of  this  character. 
The  law  of  diffuse  reflection  is  a  law  which  can  be  accurately 
applied  to  reflector  design,  but  the  possibilities  of  obtaining  va- 
ried distributions  of  light  with  diffuse  reflectors  are  exceedingly 


272  TRANSACTIONS   I.  E.  S. — PART   II 

limited.  The  users  of  reflectors  and  even  the  manufacturers  in 
many  cases,  have  quite  generally  failed  to  recognize  this. 

Referring  again  to  the  classification  of  metal  reflectors,  it 
will  be  seen  that  in  shape  there  are  four  general  classes — deep 
bowl,  shallow  bowl,  shallow  and  angle.  Deep  bowl,  shallow 
and  angle  types  of  reflectors  are  well  known.  The  shallow  bowl 
has  been  less  generally  used.  Figs  4,  6,  8  and  10  show  typical 
reflectors  of  these  four  classes.  The  deep  bowl  and  the  angle 
types  (when  the  latter  are  rightly  used)  are  much  preferable 
to  the  others  from  the  standpoint  of  protecting  the  eyes  from 
exposure  to  brilliant  lamp-filaments.  The  shallow  bowl  and 
shallow  reflectors,  however,  are  the  only  shapes  giving  a  dis- 
tributing type  of  curve  and  this  curve  has  certain  fields  of  use- 
fulness. As  a  subclassification  of  shallow  reflectors,  the  flat 
type  might  be  mentioned,  although  these  are  rapidly  becoming 
obsolete.  Flat  reflectors  usually  allow  the  lamp  filament  to  pro- 
trude too  far  below  the  lower  edge  of  the  reflector.  The  shallow 
reflectors  now  most  commonly  in  use  are  sufficiently  deep  to 
cover  the  lamp  at  least  to  the  bottom  of  the  filament. 

Distribution  of  light  is  perhaps  the  most  useful  method  by 
which  reflectors  can  be  classified.  Certainly,  in  selecting  the  re- 
flectors for  use  in  any  installation  nothing  is  more  important  than 
to  obtain  the  best  distribution  of  light  for  the  purpose.  Exten- 
sive, intensive  and  focusing  distributions  are  well-known  and 
widely  used.  The  prototype  curves  of  these  distributions  (i  e., 
the  original  curves  which  reflectors  should  be  designed  to  give) 
are  calculated  on  the  basis  of  obtaining  uniform  illumination 
when  the  light-units  are  arranged  in  squares  with  the  distance 
apart  bearing  a  definite  ratio  to  the  height  above  the  plane  of 
illumination.  These  ratios  of  distance  apart  to  height  are  2  for 
the  extensive  distribution  1%  for  the  intensive  distribution  and 
%  for  the  focusing  distribution.  Fig.  4  shows  an  illustration  of 
the  well-known  extensive  reflector  and  Fig.  5  shows  its  photo- 
metric curve;  Figs.  12  and  13  show  the  illustration  and  curve 
of  an  intensive  reflector  and  Figs.  14  and  15,  a  focusing  reflector. 
Reflectors  most  nearly  approaching  the  prototype  curves  have 
been  selected.  Reflectors  giving  these  distributions  are  available 
for  inverted  gas  and  for  nearly  all  the  various  sizes  of  tungsten 


ROLPH  I    METAL  REFLECTORS  EOR  INDUSTRIAL  LIGHTING      273 

lamps.  Extensive  and  intensive  reflectors  are  available  in  both 
aluminum  and  porcelain  enamel  finish,  although  reflectors  hav- 
ing porcelain  enamel  finish  cannot  be  designed  to  give  as  good 
an  extensive  distribution  as  reflectors  having  an  aluminum  finish. 
Focusing  reflectors  are  not  available  in  porcelain  enamel  since, 
as  stated  above,  a  focusing  distribution  cannot  be  obtained  with 
a  diffusely  reflecting  surface. 

The  distributing  type  of  photometric  curve  is  characterized 
by  high  candle-power  values  at  the  angles  of  500  to  750  from 
the  nadir  or  vertically  downward  direction.  To  obtain  high  can- 
dle-power values  at  these  angles  it  is  necessary  to  expose  the 
lamp  filament  considerably  more  than  in  the  case  of  reflectors 
giving  extensive,  intensive  and  focusing  distributions.  Conse- 
quently, while  deep  bowl  reflectors  may  be  used  to  obtain  ex- 
tensive, intensive  and  focusing  curves,  shallow  or  shallow  bowl 
reflectors  must  be  used  when  a  distributing  curve  is  desired. 
Figs.  7  and  11  show  good  distributing  curves  obtained  from  a 
shallow  porcelain  enameled  reflector  and  a  shallow  bowl  alumi- 
nized  reflector  respectively.  This  radical  difference  in  shape  of 
reflector  to  obtain  the  same  type  of  distribution  is  due  to  the 
radical  difference  in  the  character  of  the  reflection  obtained  from 
porcelain  enamel  and  aluminum,  as  explained  above.  Shallow 
reflectors  if  made  with  an  aluminum  finish  would  not  give  a  dis- 
tributing curve  unless  the  lamp  filament  were  allowed  to  project 
considerably  below  the  edge  of  the  reflector,  thus  wasting  en- 
tirely too  much  light  at  and  above  the  horizontal.  Fig.  16  illus- 
trates this,  by  showing  the  photometric  curve  of  a  shallow  type 
of  reflector  in  porcelain  enamel  and  the  curve  of  the  same  re- 
flector in  aluminum  finish.  It  will  be  seen  that  the  change  in 
finish  has  changed  the  character  of  distribution  entirely.  The 
aluminum  finished  shallow  reflector  has  little  practical  value.  It 
is  not  distributing,  while  as  a  focusing  type  it  has  too  high  candle- 
power  values  at  500  to  750  from  the  vertical,  and  does  not  protect 
the  eyes  from  the  lamp  filament  to  as  great  a  degree  as  the  regular 
types  of  focusing  reflectors.  When  a  distributing  curve  is  de- 
sired in  aluminum  finish  the  shallow  bowl  type  of  reflector  should 
be  used.  This  shape  was  developed  solely  to  obtain  a  distrib- 
uting: reflector  in  aluminum  finish. 


274  TRANSACTIONS    I.  E.  S. — PART    II 

Distributing  reflectors  have  been  more  generally  used  than 
they  should  have  been.  As  stated  above,  their  principal  charac- 
teristic is  high  candle-power  values  at  angles  of  500  to  75 °  from 
the  nadir.  To  obtain  these  high  candle-power  values  a  certain 
degree  of  eye-protection  is  sacrificed.  Consequently  they  should 
be  used  only  where  the  light  at  these  high  angles  is  of  more  im- 
portance than  the  better  eye-protection  which  would  be  obtained 
with  deep  bowl  reflectors.  There  are  certain  cases  where  light  at 
high  angles  is  of  importance.  For  example,  when  illumination 
is  required  on  many  different  vertical  or  oblique  surfaces,  high 
above  the  floor,  the  distributing  reflector  will  often  prove  the  best 
reflector  to  be  used.  It  frequently  happens,  however,  that  such 
cases  are  better  taken  care  of  by  the  proper  type  of  angle  reflec- 
tor. As  another  example,  large  stock-rooms  and  warehouses 
should  usually  be  lighted  with  distributing  reflectors.  Here  lit- 
tle actual  work  is  performed  and  exposure  of  the  lamp  filament 
is  permissible,  while  to  obtain  a  reasonably  low  cost  of  installa- 
tion, the  light-units  must  be  placed  far  apart.  When  the  distance 
apart  is  greater  than  two  and  one-half  times  the  mounting  height, 
extensive  reflectors  will  leave  dark  spots  half-way  between  light- 
units.  Distributing  reflectors  will  eliminate  these  dark  spots. 
It  must  not  be  supposed,  however,  that  distributing  reflectors  will 
give  uniform  illumination  at  wide  spacings.  Their  advantage 
lies  simply  in  the  fact  that  at  these  wide  spacings,  they  do  not 
allow  the  intensity  of  illumination  half-way  between  units  to 
drop  as  low  as  other  reflectors  do.  To  obtain  uniform  illumina- 
tion, extensive  reflectors  can  be  used  farther  apart  than  any  other 
type  of  metal  reflectors  on  the  market.  The  spacing  constant  for 
uniform  illumination  with  extensive  reflectors  is  k  =  2.  i.  e.,  dis- 
tance apart  should  be  two  times  the  height  above  the  plane  of 
illumination.  The  distributing  curve  does  not  give  uniform  illu- 
mination unless  the  spacing  is  k  =  1.6  or  less,  thus  requiring 
units  considerably  closer  together  than  when  the  extensive  distri- 
bution is  used.  This  does  not  mean  that  distributing  reflectors 
should  be  installed  at  the  spacing  k  =  1.6  or  distance  apart  1.6 
times  the  mounting  height.  When  uniform  illumination  is  de- 
sired extensive,  intensive  or  focusing  reflectors  should  be  used. 


7 


Fig.  10.— Shallow  bowl  type  of  distribut- 
ing reflector,  aluminum  finish.  (Photo- 
metric curve  shewn  in  Fig.  ii.j 


I 

^ 

w  r^ 

5r- 

v  \  j<\\ vvSO^  *°  ~^// 

//// 

\    v   >c^^.  7\  /"t^ —   ^ 

V/>) 

vxxNc^TT""*2  _^-' 

w  so» 


Fig.  ii.— Photometric  curve  of  shallow 
bowl  type  distributing  reflector,  alumi- 
num finish,  with  60- watt  clear  tungsten 
lamp  operatingat  i.i6watls  per  candle. 
I  Reflector  shown  in  Fig.  10.; 


>        c 


Fig.  12.  Intensive  aluminum  finished 
steel  reflector.  'Curve  shown  in  Fig. 
130 


Fig.  13.— Photometric  curve  of  intensive 
aluminum  finished  steel  reflector  with 
100-watt  clear  tungsten  lamp  operating 
at  1. 15  watts  per  candle.  (Reflector 
shown  in  Fig.  12.; 


I  reusing  aluminum  finished 
steel  reflector.  (Photometric  curve 
shown  in  Fig.  15.; 


Fig.  15.— Photometric  curve  of  focusing 
aluminum  finished  steel  reflector  with 
100-watt  clear  tungsten  lamp  operat- 
ing at  1. 13  watts  per  candle. 


Fig.  16. — Photometric  curves  of  shallow  dome  type  reflector, 
porcelain  enamel  and  aluminum  finished  with  250-watt 
clear  tungsten  lamp  operating  at  1.00  watt  per  candle. 


Fig.  17. — Small  angle  type  of  reflector, 
aluminum  finished.  (Photometric  curve 
shown  in  Fig.  iS.) 


li^y^'  ^fei 

\^\^C\\\^\7^r^\^^^I~/rCr~l~~L 

W  \  \-Ax^4yT\3<v/Sc  A  / / 

\\AXX^S5^/XA0 

\xxSsv\j   ^  v°  — X^\//> 

\\^\y\Cr/~ — ^K~~~C^^j<// 

/> 

vva^T^^-IIZ-6!0  "ZSA^^-a^/ 

"^ LampwithReftecTSF — 70 ^jSx^^V 

Fig.  iS. — Photometric  curve  of  small 
aluminum  finished  angle  steel  reflector 
with  25-watt  clear  tungsten  lamp  oper- 
ating at  1.1S  watts  per  caudle. 


-A"       r ./^~ 


\ 


P055IBLE 


™z< 


■  -'V 


_LJ_N  E_0  F  X1  *i5>ii 


Fig.  19. — Diagram  showing  angle  above  which  light  emitted  from 
light-units  is  likely  to  cause  eye-strain. 


ROLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      2/5 

Distributing  reflectors  are  for  wide  spacings  where  uniform  illu- 
mination is  unnecessary. 

The  fact  that  distributing  reflectors  are  less  undesirable  than 
extensive  with  very  wide  spacing  of  light-units  causes  their  use 
in  many  cases  where  closer  spacing  and  extensive  or  intensive 
reflectors  would  be  preferable.  Often  a  false  idea  of  economy 
dictates  too  small  a  number  of  outlets  and  distributing  reflectors 
are  the  only  recourse.  The  result  is  non-uniformity  of  illumina- 
tion and  lack  of  diffusion.  Shadows  are  too  dense  and  light 
comes  from  the  wrong  direction  for  many  of  the  workers.  The 
competition  among  reflector  salesmen  is  frequently  responsible 
for  this.  It  is  naturally  easiest  to  sell  the  reflector  which  can  be 
used  at  the  widest. spacings  since  the  installation  cost  is  then  the 
lowest.  Again,  the  old  fallacy  that  light  is  more  important  than 
illumination  is  responsible  for  the  use  of  many  shallow  reflectors. 
In  spite  of  the  great  spread  of  illumination  knowledge  in  the  last 
few  years,  there  is  still  a  general  impression  that  lamp  filaments 
must  be  in  plain  sight  to  obtain  the  best  results.  This  tends  to 
increase  the  sale  of  shallow  reflectors  at  the  expense  of  the  more 
desirable  deep  bowls.  A  still  more  widespread  knowledge  of 
the  principles  of  good  illumination  will  gradually  remedy  this 
condition.  Industrial  lighting  is  still  passing  through  the  stage 
which  the  general  commercial  field  passed  some  years  ago.  Shal- 
low types  of  prismatic  reflectors  are  practically  obsolete;  shallow 
opal  reflectors  are  rapidly  becoming  so,  but  shallow  metal  re- 
flectors are  still  used  in  large  quantities.  It  is  true  that  with  an 
opaque  reflector  and  in  a  field  of  such  diversified  requirements 
as  the  industrial  field,  there  is  some  need  for  a  shallow  reflector, 
but  it  must  be  admitted  that  the  use  of  this  type  is  much  more 
general  than  it  should  be. 

The  last  class  of  reflectors  given  in  the  table  under  the  classi- 
fication "Distribution  of  Light"  is  the  asymmetric  class.  Asym- 
metric distributions  are  obtained  from  angle  reflectors.  These 
reflectors  are  usually  made  up  of  symmetrical  reflector  forms 
with  the  holders  set  at  an  angle.  They  are  available  in  aluminum 
or  enamel  finish  for  all  sizes  of  tungsten  lamps  from  25-watt  or 
smaller  to  500-watt.  The  small  sizes  are  used  principally  for 
local  lighting  such  as  the  lighting  of  benches  or  particular  por- 


276  TRANSACTIONS    I.  E.  S. — PART    II 

tions  of  the  work  on  machines.  They  are  made  to  give  various 
distributions.  They  can  be  obtained  in  the  sizes  up  to  8  inches 
(20.32  cm.)  in  diameter  of  such  shape  that  the  maximum  candle- 
power  is  given  in  any  direction  desired  from  directly  downward 
to  directly  horizontal.  Most  conditions,  however,  can  be  met  by 
a  single  line  of  reflectors,  i.  e.,  one  angle  type  only  for  each  size 
lamp.  For  the  small  sizes,  used  principally  for  local  lighting,  the 
greatest  candle-power  values  should  be  between  150  and  45 °  from 
the  vertical.  For  the  larger  sizes,  the  greatest  candle-power 
values  should  be  between  35  °  and  8o°  from  the  vertical.  It 
should  be  noted  that  the  angle  at  which  the  holder  is  set  does 
not  indicate  the  direction  in  which  the  maximum  candle-power 
is  obtained.  It  is  characteristic  of  these  reflectors  that  they  give 
the  greatest  candle-power  and  greatest  flux  of  light  at  angles 
somewhat  higher  than  the  angle  at  which  the  holder  is  set.  Fig. 
17  illustrates  a  typical  small  angle  reflector  and  Fig.  18  its  pho- 
tometric curve.  A  typical  large  angle  reflector  is  shown  in  Fig. 
8  and  its  photometric  curve  in  Fig.  9.  Such  reflectors  are  not 
used  for  local  lighting  to  as  great  an  extent  as  the  small  re- 
flectors. They  are  of  value  for  general  illumination,  when  there 
are  vertical  or  oblique  surfaces  to  be  lighted  at  a  considerable 
height  above  the  floor.  They  are  also  widely  used  for  the  gen- 
eral illumination  of  spaces  in  which  it  is  difficult  to  place  sym- 
metrical reflectors  effectively.  Many  shops,  for  example,  have 
traveling  cranes  which  either  interfere  with  the  placing  of  re- 
flectors above  them  or  require  such  reflectors  to  be  placed  so  high 
as  to  lose  much  of  their  effectiveness.  In  such  cases  the  large 
angle  reflectors,  placed  on  the  upright  girders  beneath  the  crane, 
may  be  used  to  advantage. 

The  above  treatment  of  reflectors  according  to  their  various 
characteristics  does  not  include  several  important  features  which 
must  be  considered  in  selecting  a  reflector  for  any  given  service. 
In  the  classification  which  has  been  made,  eye-protection  has 
been  considered  only  slightly,  in  connection  with  the  shape  of  the 
reflector.  Efficiency,  depreciation  and  cost  have  not  been  consid- 
ered. The  classification  gives  a  broad  general  idea  of  the  char- 
acteristics of  metal  reflectors ;  but  for  an  intelligent  selection  of 
reflectors  for  any  given  installation  the  following  points  should 


R0LPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      277 

be  considered:  (1)  eye-protection;  (2)  distribution  of  light; 
(3)  efficiency;  (4)  depreciation;  (5)  cost. 

The  above  order  in  which  these  points  are  given  will,  in  many 
cases,  be  the  order  of  their  relative  importance,  but  this  will  vary 
somewhat,  depending  upon  the  character  of  the  building.  These 
five  considerations  will  be  treated  briefly  from  the  standpoint  of 
their  influence  on  choice  of  reflector. 

The  proper  protection  of  the  eyes,  as  far  as  is  known  at 
present,  involves  the  avoidance  of  three  undesirable  features, 
namely,  high  candle-power,  points  of  high  intrinsic  brilliancy  and 
extreme  contrasts  of  brilliancy  within  the  ordinary  range  of  vis- 
ion. When  work  is  actually  being  performed  the  worker  is  usu- 
ally looking  in  a  downward  direction.  Wherever  this  is  the  case, 
only  the  most  flagrant  violation  of  the  principles  of  good  lighting 
(such  as  placing  a  bare  lamp  in  front  of  the  worker  and  close 
to  the  work)  will  cause  serious  eye-strain.  However,  when  the 
worker  glances  up  from  his  work,  if  he  encounters  conditions 
causing  eye-strain,  the  eyes  are  temporarily  rendered  less  efficient 
and  it  may  be  some  minutes  after  he  looks  down  again,  before  his 
vision  is  normal.  It  is  desirable,  therefore,  in  factory  lighting  to 
minimize  the  possibility  of  eye-strain  for  all  ordinary  conditions 
of  vision.  The  first  requirement  for  this  is  to  place  light-units  as 
high  above  the  range  of  ordinary  vision  as  is  consistent  with  good 
distribution  and  diffusion  of  illumination.  In  large  rooms,  dis- 
tant light-units,  even  if  placed  high,  will  nearly  always  be  in  the 
range  of  vision  and  it  is  important  that  the  units  themselves  be 
so  designed  as  to  avoid  insofar  as  possible,  the  conditions  which 
lead  to  eye-strain.  Research  has  shown  that  high  candle-power 
ceases  to  affect  the  efficiency  of  the  eye,  when  it  is  removed  to 
25  °  from  the  direct  line  of  vision.  It  is  reasonable  to  suppose 
that  high  intrinsic  brilliancy  and  strong  brilliancy  contrasts  also 
have  little  effect  when  the  light-source  is  so  far  removed  from  the 
line  of  vision.  Assuming  that  the  direct  line  of  vision  will  not 
often  be  above  the  horizontal,  the  requirements  of  light-units  to 
obtain  best  eye-protection  are  that  the  candle-power,  intrinsic 
brilliancy  and  brilliancy  contrast  be  low  at  angles  above  25 ° 
below  the  horizontal  or  65 °  from  the  vertical.  This  requirement 
is  lenient  since  in  many  shops  the  eye  is  quite  frequently  directed 


2/8  TRANSACTIONS    I.  E.  S. — PART    II 

above  the  horizontal.  However,  the  injurious  effect  of  light 
within  this  angle  of  25 °  varies  with  the  angle.  The  greater  the 
angular  distance  of  the  light  from  the  line  of  vision  the  less  will 
be  the  injurious  effect.  The  effect,  therefore,  becomes  quite 
small  when  the  angle  is  only  slightly  less  25 °  from  the  line  of 
vision. 

The  suppression  of  candle-power  above  65 °  from  the  vertical 
is  satisfactorily  accomplished  in  the  deep  bowl  reflectors  of  the 
extensive,  intensive  and  focusing  types.  The  suppression  of 
intrinsic  brilliancy  at  these  angles  requires  the  screening  of  the 
filament  down  to  65 °  from  the  vertical.  Most  deep  bowl  reflec- 
tors do  not  meet  this  requirement  exactly ;  the  screening  angle 
varies  from  650  to  750.  Nevertheless,  deep  bowl  reflectors  are 
reasonably  satisfactory  in  this  respect  and  are,  of  course,  much 
preferable  to  the  shallow  bowl  reflectors.  The  avoidance  of 
strong  brilliancy  contrasts  above  650  from  the  vertical,  is  not 
always  easily  accomplished.  When  deep  bowl  reflectors  are  used, 
there  is  often  a  considerable  degree  of  contrast  between  the 
brilliant  interior  of  the  reflector  and  the  dark  upper  portion  of 
the  room.  Where  these  reflectors  are  used  in  rooms  having  light 
colored  ceilings  and  the  work  is  principally  on  light  colored 
goods,  this  strong  contrast  is  not  so  much  in  evidence.  In  many 
cases,  however,  there  is  practically  no  ceiling  and  frequently 
what  ceiling  there  is,  is  dark  in  color.  In  order  to  relieve  this 
contrast,  it  is  important  that  the  interior  of  the  reflector  have  as 
low  a  brilliancy  as  possible  in  the  directions  above  65°  from  the 
vertical.  If  the  character  of  the  reflection  obtained  from  an 
aluminum  surface  and  from  a  porcelain  surface  is  carefully  con- 
sidered, it  will  be  seen  that  this  has  considerable  effect  upon  the 
appearance  of  the  interior  of  the  reflector.  Aluminum  reflectors 
act  by  spread  reflection.  In  other  words  the  maximum  candle- 
power  of  the  reflected  light  is  at  the  same  angle  with  the  surface 
as  the  incident  light.  In  the  design  of  these  reflectors,  their 
contour  is  made  of  such  a  shape  that  this  feature  is  utilized  in 
directing  the  light  into  the  directions  where  maximum  candle- 
power  is  desired.  In  the  extensive,  intensive  and  focusing  dis- 
tributions, maximum  candle-power  is  desired  below  55 °  from  the 
vertical.     Consequently,   the   interior   of  the   aluminum   finished 


KOLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      2^9 

reflector  appears  brightest  when  viewed  from  below  55 °  from  the 
vertical.  Both  the  candle-power  and  the  intrinsic  brilliancy  of 
the  interior  surface  are  comparatively  low  at  650  and  above. 
With  the  porcelain  enameled  reflector,  conditions  are  somewhat 
different.  The  reflection  is  diffuse  and  the  maximum  candle- 
power  from  every  point  on  the  surface  is  in  a  direction  normal  to 
the  surface.  With  diffuse  reflection  much  of  the  light  at  each 
reflection  is  directed  back  into  the  reflector  and  strikes  the  sur- 
face again.  The  consequence  is  that  the  entire  interior  surface 
of  the  reflector  has  approximately  the  same  degree  of  brightness. 
In  looking  at  such  a  reflector  at  angles  above  65 °  from  the  ver- 
tical, one  sees  an  interior  surface  of  the  same  brightness  as  would 
be  seen  from  directly  underneath.  The  candle-power  is  low  at 
these  high  angles  because  the  edge  of  the  reflector  cuts  off  the 
light  from  a  large  part  of  the  opposite  side  of  the  reflector.  The 
intrinsic  brilliancy,  however,  and  the  degree  of  brilliancy  contrast 
are  considerably  higher  than  with  an  aluminum  finish  reflector  of 
the  same  shape. 

These  considerations  indicate  that  from  the  standpoint  of  eye- 
protection,  the  most  desirable  reflector  is  the  deep  bowl  in  alu- 
minum finish ;  second  choice  is  the  deep  bowl  in  porcelain  enamel 
finish,  while  least  desirable  from  this  standpoint  are  the  shallow 
bowl  and  shallow  types  of  reflector. 

The  question  of  distribution  of  light  has  been  covered  above 
in  a  general  way.  It  is  not  necessary  here  to  describe  in  detail 
the  characteristics  of  the  extensive,  intensive,  focusing  and  dis- 
tributing photometric  curves.  These  distributions  are  well  known 
and  their  relative  merits  and  uses  have  been  treated  before.  It 
is  important  that  reflectors  should  be  selected  for  a  given  service 
by  their  distribution  of  light.  There  is  too  great  a  tendency  at 
present  to  consider  shape  of  reflector  as  important  in  selec- 
tion. Having  settled  the  question  of  eye-protection  in  any  given 
installation,  shape  should  be  entirely  disregarded  and  distribution 
considered.  For  example,  the  distributing  type  of  photometric 
curve  is  obtained  from  two  reflectors  of  entirely  different  shape, 
one  in  aluminium  finish  and  one  in  enamel  finish.  The  shape  of 
these  two  reflectors  should  be  neglected  in  choosing  between  them 
and  the  choice  made  on  the  basis  of  their  other  characteristics. 
6 


28o  TRANSACTIONS   I.  Z.  S. — PART    II 

In  comparing  reflectors  on  the  basis  of  efficiency,  it  is  im- 
portant to  consider  efficiency  in  the  sense  of  illumination  obtained 
for  energy  expended.  The  total  output  of  a  reflector  is  no  meas- 
ure of  its  illumination  efficiency.  If  this  were  the  case,  a  bare 
lamp  would  be  more  efficient  that  a  lamp  and  reflector.  It  is 
obvious  that  a  shallow  reflector  will  have  a  greater  total  output 
of  light  than  a  corresponding  deep  reflector,  since  with  the  shal- 
low reflector  much  less  of  the  light  from  the  lamp  strikes  the 
reflector  and  there  is,  consequently,  less  absorption  at  the  re- 
flecting surface.  This  greater  total  output  with  shallow  reflectors 
does  not  mean  that  a  higher  percentage  of  the  flux  of  the  lamp 
will  be  useful.  Only  in  exceptional  cases  is  this  greater  total 
output  a  distinct  advantage.  In  comparing  the  efficiency  of  re- 
flectors in  any  given  installation,  an  illumination  test  is  the  most 
satisfactory  method.  Certain  comparisons  from  the  photometric 
curve,  however,  may  be  made  to  advantage  when  an  illumination 
test  is  impractical.  The  light-flux  in  the  zone  o°  to  6o°,  or 
below  6o°  from  the  vertical,  is  used  considerably  in  comparing 
the  effective  flux  of  reflectors.  This  zone  includes  on  the  aver- 
age, the  light  striking  the  plane  of  illumination  directly,  in  large 
rooms.  In  medium-sized  rooms,  the  angle  including  the  directly 
effective  flux  is  somewhat  lower.  Flux  in  the  zone  o°  —  6o°  is 
a  much  better  basis  of  comparison  of  the  efficiency  of  reflectors, 
than  total  output  or  flux  below  the  horizontal. 

The  specific  reflecting  efficiencies  of  aluminum  finish  and  porce- 
lain enamel  finish  are  very  nearly  the  same.  Porcelain  enamel 
will  usually  run  somewhat  higher.  The  actual  efficiency  obtained 
from  any  given  reflector,  however,  depends  much  upon  the  shape 
of  the  reflector.  With  shallow  reflectors,  porcelain  enamel  is 
considerably  more  efficient  than  aluminum.  With  the  deep  types, 
however,  the  diffuse  character  of  the  reflection  from  porcelain 
enamel  tends  to  cause  much  of  the  light  to  be  reflected  back  and 
forth  within  the  reflector.  This  increases  the  amount  of  light 
absorbed,  and,  as  a  consequence,  in  the  deep  bowl  reflectors, 
aluminum  finish  is  more  efficient  than  porcelain  enamel. 

The  deterioration  of  reflectors,  due  to  the  collection  of  dust,  is 
an  important  consideration  in  industrial  lighting.  Dust  and 
dirt  are  much  more  prevalent  than  in  the  general  lighting  field 


R0LPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      28l 

and  consequently  more  apt  to  reduce  the  efficiency  of  the  lighting 
system.  Among  metal  reflectors  the  principal  difference  in  this 
respect  is  due  to  difference  in  finish.  Reflectors  with  a  smooth 
interior  surface  such  as  porcelain  enamel  are  naturally  more 
easily  cleaned  than  those  with  a  rough  surface  such  as  the  usual 
types  of  matte  aluminum.  Furthermore,  porcelain  enamel  re- 
flectors after  cleaning  regain  their  initial  reflecting  efficiency, 
while  the  usual  types  of  aluminum  finish,  after  once  becoming 
dirty,  can  never  be  cleaned  sufficiently  to  regain  their  original 
condition.  They  always  show  a  permanent  deterioration  in  re- 
flecting efficiency  of  5  per  cent,  or  more,  the  exact  percentage 
depending  upon  the  character  of  the  dirt.  It  is  unfortunate  that 
this  is  the  case  in  view  of  the  other  advantages  possessed  by 
aluminum  finished  reflectors.  The  search  for  a  method  of  over- 
coming this  permanent  deterioration  and  difficulty  of  cleaning 
has  not  been  entirely  without  success.  A  finish  has  recently  been 
developed  which  will  undoubtedly  do  much  to  increase  the  popu- 
larity of  aluminum.  This  consists  of  an  interior  finish  compris- 
ing three  distinct  coats  of  different  material  instead  of  the  one 
coat  of  aluminum  usually  employed.  The  lower  ground  coat 
consists  of  a  white  material  which  is  practically  impervious  to 
moisture.  This  serves  the  double  purpose  of  protecting  the  re- 
flector from  rust  and  slightly  increases  the  efficiency  since  some 
light  usually  penetrates  the  aluminum  finish  in  places.  The  sec- 
ond coat  is  the  ordinary  aluminum  finish.  The  third  consists  of 
a  washable  lacquer.  This  forms  a  smooth  hard  surface  which  is 
easily  cleaned,  and  when  cleaned,  restores  the  surface  to  its  initial 
efficiency.  This  coat  is,  of  course,  transparent  and  the  absorp- 
tion is  so  low  that  the  decrease  in  efficiency  due  to  its  use  is  much 
less  than  the  permanent  deterioration  of  the  ordinary  alumi- 
num finish  after  its  first  cleaning.  In  addition  to  the  three  inside 
coats,  the  outside  of  the  reflector  is  finished  with  the  same  ground 
coat  as  the  inner  surface,  after  which  the  usual  paint  enamel  is  ap- 
plied. The  reflector,  therefore,  has  5  coats  instead  of  the  2 
ordinarily  used.  This  finish  bids  fair  to  overcome  the  last 
objection  to  the  use  of  aluminum  finished  reflectors.  Aluminum 
finish  will  undoubtedly  become  more  popular  than  ever  before 
and  gradually  replace  the  less  desirable  porcelain  enamel. 


282  TRANSACTIONS   I.  E.  S. — PART    II 

The  comparative  cost  of  various  reflectors  is  a  subject  which 
must  be  considered  for  the  individual  case  rather  than  in  any  gen- 
eral treatment  of  the  subject.  Reflectors  vary  in  cost  due  to  dif- 
ferences in  finish  more  than  due  to  differences  in  shape.  Of  the 
finishes  in  common  use,  the  aluminum  is  somewhat  lower  in  cost 
than  porcelain  enamel.  Initial  cost  is  worthy  of  very  little  atten- 
tion, however.  The  value  of  obtaining  the  best  illumination  re- 
sults far  outweighs  such  considerations  as  cost  of  reflector  and 
cost  of  installation.  Take  for  example,  the  lighting  of  a  single 
machine.  The  best  reflector  for  the  purpose  may  cost  80  cents. 
The  machine  is  worth  several  hundred  or  very  likely,  several 
thousand  dollars.  Considering  the  reflector  as  a  necessary  part 
of  the  equipment  of  the  machine  it  is  apparent  that  its  cost  is  in- 
significant. Again,  consider  the  value  of  the  workman's  time; 
25  cents  an  hour  is  a  low  price,  but  even  at  that  figure  the  loss  of 
only  a  few  hours  due  to  poor  eye-protection  or  poor  illumination, 
would  offset  the  entire  cost  of  the  best  reflector  obtainable.  Or, 
take  the  value  of  the  work  performed.  An  hour's  work  of  a  sin- 
gle workman  may  be  actually  worth  several  dollars  to  the  com- 
pany. Yet  one  mistake  due  to  poor  eye-protection  or  poor  illu- 
mination may  require  many  hours'  work  to  rectify  it.  These 
comparisons  are  sufficient  to  show  that  reflectors  should  not  be 
selected  by  cost  but  rather  by  the  results  which  they  produce. 
Illuminating  engineering  has  advanced  to  such  a  position  that  the 
best  lighting  system  which  a  factory  can  install  is  usually  worth 
many  times  what  it  actually  costs. 

From  the  above  treatment  of  the  factors  influencing  the  choice 
of  reflector,  the  following  generalizations  may  be  made. 

1.  Aluminum   finished   reflectors   are  preferable   to   porcelain 

enamel  from  the  standpoints  of 

a.  Variety  of  distribution  obtainable. 

b.  Protection  of  the  eyes. 

c.  Low  cost. 

2.  Porcelain    enamel    reflectors    are   preferable   to    aluminum 

from  the  standpoint  of  ease  of  maintenance  and  lack  of 
permanent  deterioration  except  when  the  aluminum  finish 
is  protected  by  a  smooth  transparent  lacquer;  in  that  case 


ROLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      283 

the  two  finishes  are  probably  equally  good,  for  most 
classes  of  service. 

3.  Deep  bowl   reflectors  are  preferable  to  shallow  bowl  and 

shallow  reflectors  from  the  standpoint  of, 

a.  Protection  of  the  eyes. 

b.  Variety  and  usefulness  of  distributions  available. 

(Deep  bowl  reflectors  are  available  giving  ex- 
tensive, intensive  and  focusing  distributions. 
The  shallower  reflectors  give  only  a  distrib- 
uting type  of  curve.) 

4.  Extensive,   intensive  and  focusing  photometric  curves  are 

preferable  to  the  distributing  curve  for  the  majority  of 
cases  of  general  lighting.  The  distributing  curve  should 
be  reserved  for  special  cases  such  as  warehouses,  stock- 
rooms, etc.,  or  work-rooms  in  which  light  is  required  on 
high  vertical  surfaces. 

5.  Initial  cost  of  reflectors  is  of  minor  importance.     It  should 

usually  be  disregarded,  except  in  comparing  reflectors 
which  are  equally  good  in  other  respects. 

DISCUSSION. 

Mr.  H.  Calvert  :  It  is  certainly  true  that  in  the  industrial 
field  there  are  many  crimes  committed  against  the  conservation 
of  vision, — bare  lamps  suspended  in  the  line  of  vision,  improper 
reflectors,  and  lamps  badly  located.  Comparing  industrial  light- 
ing with  the  lighting  of  stores,  the  difference  shows  up  very 
much  to  the  disadvantage  of  the  factories  and  workshops,  and 
I  think  the  reason  is  not  very  hard  to  find.  The  storekeeper 
naturally  wants  to  attract  the  public  into  his  store  and  to  do 
this  he  is  generally  willing  to  spend  a  certain  amount  to  improve 
his  illumination,  thinking  in  doing  so  that  it  will  increase  his 
profits.  The  factory  manager,  on  the  other  hand,  does  not  come 
in  contact  with  the  public ;  the  public  does  not  enter  his  factories, 
and  he  is  very  apt  to  consider  that  any -money  spent  on  such 
illumination  means  a  decrease  in  his  profits. 

There  is  no  question  that  great  improvements  can  be  made 
by  the  reasonable  use  of  metallic  reflectors.     I  have  in  mind  a 


284  TRANSACTIONS    I.  %  S. — PART    II 

large  weave  room  which  was  originally  illuminated  by  means 
of  86  series  arc  lamps  with  clear  globes.  These  were  replaced 
by  171  100-watt  tungsten  lamps  with  shallow  bowl  reflectors. 
The  result  was  a  great  improvement  in  the  lighting,  greater 
uniformity,  and  a  decrease  of  approximately  40  per  cent,  in  the 
energy  consumed. 

This  installation  brings  to  mind  an  interesting  psychological 
study.  The  same  mill  had  a  similar  weave  room  which  was 
lighted  practically  the  same  way — the  same  illumination,  but 
in  the  first  room  of  which  I  spoke  the  lamps  are  practically  ex- 
posed. In  the  second  room  a  deeper  shade  is  used,  so  that  the 
lamps  are  entirely  hidden  from  view.  Now  the  operators  in  the 
second  room  are  firmly  of  the  opinion  that  they  do  not  get  as 
much  light  as  the  operators  in  the  first  room,  simply  because 
they  cannot  see  the  sources  of  light. 

Mr.  C.  O.  Bond:  I  would  wish  only  to  call  attention  to  one 
installation  of  deep  bowl  metal  reflectors,  which  I  think  would 
appeal  to  anyone  seeing  it,  that  is  along  the  line  of  the  under- 
ground platform  at  the  West  Philadelphia  station  of  the  Penn- 
sylvania Railroad  Company.  As  one  goes  down  to  take  the 
train  to  Baltimore,  or  going  South,  there  is  quite  a  long  platform. 
In  going  up  or  down  such  platforms  one  receives  a  glare  of 
light  in  the  eye;  and  the  approaching  engineer  is  so  blinded  as 
he  comes  down  the  track  that  he  almost  finds  himself  in  the 
position  of  being  unable  to  tell  whether  he  is  running  into  danger 
or  not.  The  use  of  deep  bowl  reflectors  has  solved  the  problem 
in  this  particular  case  very  well  indeed  and  I  think  it  is  worth 
while  for  anyone  to  study  that  installation. 

I  should  like  to  ask  Mr.  Rolph  if  the  local  reflection  in  the 
deep  bowl  reflectors  has  any  influence  whatever  on  the  length 
of  original  life  of  the  filament.  I  was  wondering  whether  there 
is  a  similar  effect  found  by  frosting  the  bulbs  of  electric  lamps, 
in  which  I  think  the  shorter  life  is  somewhat  due  to  the  increased 
temperature.    I  ask  for  that  information. 

Prof.  George  A.  Hoadley:  Mr.  Calvert  has  called  attention 
to  the  fact  that  in  many  factories  there  are  high  ceilings,  and  I  am 


ROLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      285 

pretty  sure  that  anyone  present  can  appreciate  the  difficulties  in 
lighting  such  places. 

Mr.  Rolph  has  spoken  of  the  metal  reflectors,  and  it  occurred 
to  me  that  in  their  use  we  get  the  type  of  lamp  known  as  the 
direct  type.  Now  is  there  any  particular  portion  of  the  light 
that  comes  from  these  reflectors  that  is  reflected  on  the  walls, 
thus  giving  a  bad  effect  in  the  illumination  of  the  room  ?  I  have 
thought  of  that  as  one  of  the  things  which  might  be  taken  up. 

Another  point  is,  that  it  seems  to  me  it  would  add  very  much 
to  our  information  if  we  could  have  two  illustrations — one  show- 
ing the  ordinary  daylight,  and  the  other  which  would  give  what 
we  might  term  artistic  illumination.  Then  we  would  be  able  to 
make  a  contrast  between  the  two.  But  if  we  have  only  one,  or 
only  daylight,  while  we  get  sufficient  information  in  connec- 
tion with  the  lamp,  it  does  not  seem  to  me  that  it  gives  all  the 
information  we  ought  to  get. 

Prof.  Arthur  J.  Rowland:  Some  one  ought  to  write  the 
history  of  shades.  I  presume  that  on  sources  of  light  being 
open  flames  they  were  originally  used  for  either  of  two  purposes : 
to  cut  light  off  from  a  particular  direction  where  it  was  not 
desired,  or  to  serve  to  reduce  the  risk  of  setting  fire  to 
objects  in  the  vicinity.  I  incline  to  the  opinion  that  most  shades 
used  on  electric  lamps,  prior  to  the  day  of  tungsten  lamps,  were 
there  because  their  use  was  in  accord  with  common  practicable 
gas  burners ;  not  to  produce  any  special  redistribution  of  light, 
or  to  soften  the  brightness  of  a  small  area  light  source.  Then 
again  one  feels  that  a  bare  light  is  uncouth  and  crude. 

When  it  comes  to  shades  planned  to  redistribute  light,  it  is 
easy  to  think  there  are  great  differences  where  none  exist.  Take 
a  simple,  opaque,  cone  shade  with  a  white  diffusing  surface;  the 
redistribution  produced  is  practically  independent  of  the  angle 
of  the  cone  or  the  way  it  is  placed  with  reference  to  the  lamp ; 
except  that  more  or  less  light  is  entirely  cut  off  as  the  light  is 
pushed  further  into  the  cone  or  partly  withdrawn.  I  am  wonder- 
ing how  much  these  more  modern  opaque  shades  differ  from  the 
simpler  and  older  forms  in  the  light  distribution  they  are  able 
to  produce. 


286  TRANSACTIONS    I.  E.  S. — PART    II 

In  Mr.  Rolph's  paper  the  word  efficiency  has  been  used  quite 
a  little,  but  nothing  has  been  said  about  how  much  the  efficiency 
is;  that  is  what  per  cent,  of  the  total  light  the  bare  lamp  would 
pour  out,  is  still  poured  out  from  the  lamp  with  any  of  these 
shades  over  it.  It  seems  curious  to  me  how  little  manufacturers 
care  to  give  that  information.  Efficiency  ought  to  be  as  im- 
portant as  candle-power  distribution,  for  this  last  has  great  im- 
portance only  where  a  relatively  small  number  of  sources  is  in 
use. 

I  wonder  whether  some  of  the  first  intelligent  efforts  to  pro- 
duce efficient  shades  for  ordinary  lights  giving  a  desirable  light 
distribution  were  not  made  right  here  in  Philadelphia.  Prior  to 
the  time  when  the  names  Holophane-D'Olier  were  coupled  as 
describing  a  certain  line  of  opaque  shades,  a  large  part  of  the 
line  was  made  by  the  D'Oliers  here  in  our  town.  At  the  begin- 
ning of  that  work,  if  my  recollection  serves  me  correctly  Prof. 
Edwin  Houston,  so  well  known  in  Philadelphia,  made  the  de- 
signs for  the  shades,  the  metal  used  being  aluminum.  Shortly 
after  this  time  the  factory  superintendent  struck  trouble.  The 
aluminum  shade  with  its  polished  reflecting  surface  reflected  the 
light  all  right  but  the  "streaky"  illumination  was  produced.  Some- 
thing had  to  be  done.  The  superintendent  had  to  devise  some 
remedy.  In  his  difficulty  he  came  out,  for  some  reason  or  other, 
to  see  me  at  Drexel  Institute.  He  stated  the  case  and  asked  me 
to  suggest  some  process  by  the  use  of  which  the  inside  of  those 
aluminum  shades  could  be  given  some  sort  of  a  matt  surface. 
I  had  to  tell  him  I  was  unable  to  help,  since  the  question  was  a 
chemical  one.  I  took  him  to  Prof.  Henwood,  head  of  the  depart- 
ment of  chemistry  who  said  at  once,  "Wash  them  with  a  little 
caustic  soda."  "Oh,"  said  my  friend,  "Will  anything  as  simple  as 
that  really  do  it?"  "Yes,"  said  Prof.  Henwood,  "Nothing  more  is 
necessary — try  it."  After  that  they  made  their  whole  product 
that  way. 

I  had  hoped  that  this  evening  the  discussion  would  leave  the 
subject  of  shades  and  consider  the  broader  subject  of  industrial 
lighting.  There  is  one  very  important  great  detail  of  industrial 
lighting  which  seems  to  have  been  given  but  little  attention  by 
anyone ;  that  is,  the  accumulation  of  data  which  will  give  the  an- 


ROLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      287 

swer  to  this  question,  "Is  it  or  is  it  not  worth  while  to  light  rooms 
and  machinery  correctly  and  well  ?" 

About  a  week  ago  I  learned  from  a  man  who  has  made  an 
investigation  to  determine  the  effect  of  artificial  lighting  on  out- 
put and  on  the  operator.  That  nervous  strain  as  indicated  by 
the  presence  of  nervous  headaches,  could  be  eliminated,  in  the 
case  of  girls  working  in  the  printing  industries,  by  using  appro- 
priately shaded  lamps  and  requiring  the  operators  to  wear  eye- 
shades.  He  added  that  the  time  required  on  adjusting,  making 
repairs  to  machinery,  etc.,  could  be  very  considerably  diminished 
by  a  permanent  installation  of  lights  at  the  places  where  adjust- 
ments and  repairs  were  made.  The  results  he  said  were  almost 
too  good  to  believe,  and  that  he  would  not  tell  anyone  how  good 
they  were  because  he  feared  the  facts  might  be  discredited.  He 
had  not  reached  the  point  of  considering  quality  of  output,  or 
safety  of  employees,  when  the  investigation  was  discontinued. 
Surely  some  day  some  one  will  conduct  an  elaborate  investigation 
taking  up  and  determining  not  only  the  effect  on  the  output,  as 
to  both  quality  and  quantity,  when  good  artificial  lighting  is 
provided  and  whether  its  expense  is  warranted ;  but  also  how  a 
worker  stands  his  day's  work  and  how  much  increased  safety  is 
secured  to  him  by  the  use  of  good  lighting. 

Such  questions  are  as  important  as  any  which  can  be  consid- 
ered in  connection  with  industrial  lighting.  The  kind  of  lights 
used,  their  arrangement,  the  kind  of  shades  put  on  them,  are 
insignificant  matters  compared  with  the  money  value  of  good 
light  to  the  industries.  This  will  have  to  be  determined  somehow 
if  industrial  lighting  is  to  come  into  its  own. 

Mr.  J.  W.  Lee:  I  should  like  to  ask  Mr.  Rolph's  opinion 
regarding  glass  reflectors  and  metal  reflectors  for  industrial  light- 
ing. It  appears  to  me  that  in  large  rooms,  where  the  ceilings 
and  the  walls  are  white,  and  the  lamps  are  hung  high,  that  glass 
reflectors  have  a  particularly  pleasing  effect — the  cheerful  aspect 
of  the  light  ceilings  and  light  walls  have  such  an  effect,  to  my 
mind.  ■  Is  there  any  real  objection  to  a  glass  reflector — which  is 
more  efficient  than  either  the  aluminum  or  metal  reflectors — 
being  used  for  industrial  lighting? 


288  TRANSACTIONS   I.  E.  S. — PART    II 

Mr.  T.  W.  Rolph  (In  reply)  :  In  regard  to  the 
point  that  Mr.  Bond  brought  out  on  the  effect  of 
multiple  reflections — I  do  not  believe  that  I  have  any- 
more information  that  he  has  on  that  point,  with  the 
possible  exception  of  the  heating  effect.  Of  course  this  multiple 
reflection  is  a  reflection  of  heat  as  well  as  light,  and  I  can 
say  that  the  heating  effect  has  no  effect  upon  the  life  of 
the  lamps.  Careful  temperature  tests  which  have  been  made 
show  that  the  temperature  obtained  in  all  types  of  deep  bowl 
reflectors — aluminum  or  porcelain  enamel — is  not  sufficient  to 
affect  the  life  at  all.  It  is  reasonable  to  suppose  that  the  de- 
crease in  the  effective  light  of  the  lamps  would  be  greater  for 
porcelain  enamel  deep  bowl  reflectors,  on  account  of  the  multiple 
reflection  back  through  the  bulb;  and  when  the  bulb  begins  to 
blacken  the  absorption  will  increase.  As  Mr.  Bond  pointed  out, 
it  is  exactly  the  same  as  with  frosted  lamps.  With  frosted  lamps 
the  actual  life  of  the  lamp  is  just  as  great  as  with  the  clear 
lamps,  but  the  effective  light,  to  80  per  cent,  of  candle-power  is 
only  about  half  as  great — simply  due  to  the  cross-reflection  in 
the  bulb. 

Professor  Hoadley  brought  out  the  point  of  the  reflection 
from  walls  with  metal  reflectors.  Of  course  with  small  installa- 
tions there  is  considerable  reflection  from  the  walls  and  the  actual 
diffusion  of  course  is  bad.  Diffusion  is  obtained  by  light  from 
distant  units  in  large  installations.  With  metal  reflectors  there 
is  not  as  great  diffusion,  as  a  general  rule,  as  with  glass  reflectors, 
but  in  industrial  lighting  usually  metal  reflectors  can  be  so  placed 
that  the  direction  of  the  light  will  be  satisfactory  for  direct  light, 
so  that  great  diffusion  is  not  necessary. 

The  question  of  glass  versus  steel  is  one  that  ought  to  be 
discussed  at  considerable  length  if  discussed  at  all.  I  will 
barely  touch  on  it,  on  the  point  of  efficiency.  The  most  efficient 
types  of  glass  reflectors  are  more  efficient  than  steel  when  the 
ceilings  are  very  light  in  color.  When  ceilings  are  dark  in  color 
the  efficient  types  of  steel  reflectors  give  a  higher  efficiency 
than  most  types  of  glass  reflectors.  There  is  a  question  of 
depreciation  to  be  considered.  Of  course  when  bare  lamps  and 
glass  reflectors  are  used  to  give  a  higher  efficiency  in  illumination. 


ROLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      289 

the  efficiency  which  they  give  is  obtained  by  reflection  from 
the  ceiling  to  a  certain  extent,  and  naturally  the  depreciation 
would  be  a  little  greater  than  with  steel  reflectors.  There  are 
undoubtedly  many  cases  where  glass  reflectors  are  better  than 
steel  reflectors  in  industrial  lighting.  But  in  the  great  major- 
ity of  cases  it  seems  that  steel  would  be  more  satisfactory. 
Glass  reflectors  are  appropriate  in  rooms  where  the  conditions 
of  dirt  are  not  bad,  and  in  general,  in  light-colored  rooms, 
where  the  work  is  clean,  glass  reflectors  are  quite  satisfactory. 
As  I  say,  however,  that  is  too  big  a  question  to  discuss  here  in 
detail.  I  did  not  quite  get  Professor  Rowland's  point,  on  the 
diffused  reflection — on  the  character  of  distribution  you  get  with 
a  reflector  when  the  reflector  is  used  more  as  a  shade. 

Prof.  A.  J.  Rowland:  No  matter  what  the  angle  of  the  re- 
flector, or  what  its  size,  the  redistribution  of  light  is  essentially 
the  same. 

Mr.  T.  W.  Rolph  :  Yes,  that  is  true,  with  diffusing  reflec- 
tors ;  the  general  distribution  of  light  from  a  diffuse  reflector 
is  very  nearly  the  same,  no  matter  what  the  shape.  The  shape 
of  the  reflector  has  some  bearing  on  efficiency — not  very  much 
on  the  distribution  of  light — and  the  extensive  curves  obtained 
with  porcelain  enamel  reflectors  are  not  as  good — not  as  wide, 
truly  typical  extensive  curves — as  those  obtained  with  aluminum 
reflectors. 

Professor  Rowland  asked  about  the  efficiency  of  reflectors. 
Deep  bowl  aluminum  reflectors  give  a  total  output  of  55  per  cent. 
to  65  per  cent,  of  the  total  flux  of  the  lamp.  Deep  bowl  enamel 
will  run  from  50  per  cent,  to  65  per  cent.  Some  of  the  deep  bowl 
in  enamel  finish  will  run  just  as  high  as  in  aluminum  finish,  but 
those  are  reflectors  in  which  the  widest  part  does  not  come  down 
quite  as  far  as  the  corresponding  aluminum  reflector.  The  deep 
bowl  reflectors,  which  come  down  well  over  the  filament  protect- 
ing the  eyes  well,  are  a  little  bit  under  the  aluminum  finish  in  effi- 
ciency. In  the  shallow  type  of  reflector  less  light  strikes  the 
reflector  and  consequently  shallow  reflectors  will  vary  from  70 
per  cent,  to  85  per  cent,  depending  for  the  efficiency  of  reflection 
upon  the  character  of  the  material  used. 


29O  TRANSACTIONS    I.  £.  S. — PART    II 

There  is  a  great  deal  to  be  learned,  as  Prof.  Rowland  pointed 
out,  in  regard  to  the  effect  of  good  illumination  upon  the  effi- 
ciency of  the  work  in  industrial  lighting.  There  are  many  in- 
vestigations which  ought  to  be  made  very  much  along  the  same 
lines  as  the  investigations  which  have  been  made  on  motion 
study  and  investigations  along  such  lines  as  that  will  show 
the  efficiency  of  the  workers  under  various  systems  of  illumina- 
tion. I  have  no  doubt  when  we  get  data  of  that  character  it 
is  simply  going  to  be  astounding,  in  showing  the  increase  in  out- 
put and  decrease  in  accidents,  and  such  effects  of  good  illumina- 
tion. Undoubtedly  such  investigations  will  be  made  in  the  near 
future —  a  great  many  of  them — because  it  is  too  big  a  commer- 
cial question  to  leave  alone  very  long. 

Mr.  W.  F.  LITTLE  (Communicated)  :  Mr.  Rolph  has  given 
very  clearly  his  views  on  the  ideal  reflector  for  industrial  light- 
ing and  has  put  on  record  his  opinions  as  to  the  best  material 
and  shape  as  well  as  on  the  proper  methods  of  installations. 

He  refers  also  to  the  very  noteworthy  development  of  a  more 
durable  aluminum  finished  reflector.  In  the  past  it  has  not  been 
exceptional  nor  even  unusual  to  find  a  reflector  of  this  type  which 
has  shown  a  depreciation  in  reflecting  power  of  15  to  20  per 
cent,  in  a  short  time,  even  though  it  had  received  more  than 
ordinary  care.  Once  the  lustre  is  gone,  an  aluminum  finished 
reflector  of  the  ordinary  type  has  permanently  lost  much  of  its 
efficiency.  If  the  white  protective  lacquer  referred  to  by  Mr. 
Rolph  reduces  the  efficiency  but  5  per  cent.,  and  prevents  further 
depreciation,  its  use  marks  an  important  improvement. 

Mr.  Rolph  will  probably  find  many  who  feel  that  the  flat  or 
shallow  type  of  porcelain  enamel  reflector  is  not  such  a  back 
number.  In  a  great  deal  of  industrial  lighting  where  opaque 
reflectors  are  suitable,  the  conditions  are  such  as  to  require 
general  illumination  at  an  angle  greater  than  65  degrees  from  the 
vertical.  Furthermore,  by  using  a  distributing  reflector  of  some- 
what higher  efficiency,  throwing  some  illumination  nearer  the 
horizontal,  the  contrast  between  the  reflector  and  background 
may  perhaps  be  lessened  without  a  substantial  loss  in  the  illumi- 
nation on  the  working  plane,  thus  producing  effects  more  agree- 


ROLPH  :    METAL  REFLECTORS  FOR  INDUSTRIAL  LIGHTING      29I 

able  and  equally  free  from  eye  strain.  Possibly  in  a  majority 
of  cases  this  illumination  near  the  horizontal  will  not  be  wasted. 

It  is  further  implied  by  Mr.  Rolph  that  for  most  installations 
the  bowl  type  aluminum  reflector  is  preferable  to  the  porcelain 
enamel  on  account  of  the  fact  that  aluminum  produces  an 
"irregular  reflection"  while  enamel  produces  a  "diffuse  reflec- 
tion." Also  that  the  light  between  65  degrees  and  horizontal 
decreases  the  ability  to  see,  and  that  the  strain  increases  with  the 
angle.  Further,  that  the  specific  intensity  of  the  reflecting  sur- 
face will  be  greater  in  the  enamel  bowl  type  reflector  than  in 
the  aluminum  reflector  of  the  same  type  because  of  the  reasons 
indicated. 

In  comparing  the  two  types  of  reflector  it  should  be  noted, 
first,  that  the  degree  of  diffuse  reflection  secured  from  porcelain 
enamel  is  largely  dependent  on  the  quality  of  the  enamel  surface. 
It  is  quite  possible  to  secure  enameled  reflectors  which  will  con- 
trol light  almost  as  accurately  as  aluminum  reflectors.  Enamel 
of  this  quality  will  of  course  be  of  a  higher  order  of  efficiency. 
Second,  while  it  is  true  that  the  light  between  65  degrees  and 
the  horizontal  decreases  somewhat,  the  ability  to  see,  nevertheless 
experiments  have  indicated  that  the  amount  of  the  decrease  has 
been  greatly  overestimated  and  that  it  is  not  nearly  as  important 
as  some  observers  have  maintained. 

Mr.  T.  W.  Rolph  (Communicated)  :  Mr.  Little's  statement 
in  regard  to  enameled  surfaces  which  do  not  give  a  high  degree 
of  diffuse  reflection  is  very  interesting.  If  such  surfaces  can  be 
used  for  reflectors  and  can  control  the  light  as  accurately  as 
aluminum  finished  reflectors,  they  have  a  wide  field  of  usefulness. 
A  reflecting  surface,  for  use  on  metal  reflectors,  which  will  con- 
trol light  accurately  and  which  can  be  as  easily  cleaned  as  por- 
celain enamel  and  which  does  not  give  the  disagreeable  streaked 
effect  of  polished  metal,  will  be  a  very  valuable  addition  to  the 
materials  in  use  at  present. 


292  TRANSACTIONS    I.  E.  S. — PART    II 

VISION  AS  INFLUENCED  BY  THE  BRIGHTNESS  OF 
SURROUNDINGS.* 


BY  PERCY  W.  COBB. 


The  often-repeated  remark  that  the  number  of  foot-candles 
upon  the  work  is  not  an  adequate  indication  of  the  virtues  of  an 
illumination  system,  even  solely  with  respect  to  that  particular 
work ;  the  recounting  of  experience  in  corroboration  of  this  fact, 
and  the  experimental  attempts  that  are  being  made  to  ascertain 
the  reason,  and  give  quantitative  expression  to  it,  are  conditions 
which  make  a  very  limited  amount  of  introductory  matter  ade- 
quate to  the  present  paper. 

This  paper  constitutes  an  attempt  to  answer  the  question: 
"How  is  vision  influenced  by  the  bright,  visible  surroundings  of  its 
particular  object?"  The  experimental  method  by  which  the  work 
here  outlined  was  done  is  briefly  thus :  An  observer  sitting  in  a 
dark  room  faces  a  bright  surface  of  small  dimensions  trans- 
illuminated  from  the  next  room.  At  a  given  moment  the  bright 
spot  thus  seen  is  replaced  for  a  short  time  by  a  field  of  black 
and  white  lines  of  the  same  outside  dimensions  and  the  same 
average  brightness.  By  repeating  the  experiment  with  lines  of 
various  widths  the  exact  width  of  the  lines  can  be  determined 
which  is  necessary  in  order  that  they  may  be  just  visible. 

Similarly,  instead  of  using  a  lined  surface,  the  original  blank 
surface  can  be  replaced  by  a  field  of  the  same  brightness,  except 
that  one  half  of  it  is  increased  or  diminished  by  a  small  fraction 
of  its  intensity.  In  this  way  can  be  determined  the  exact  differ- 
ence necessary  in  order  that  it  may  be  seen  as  a  difference.  As 
both  of  these  quantities  (smallest  visible  detail  and  smallest  vis- 
ible difference)  vary  with  variations  in  brightness,  the  determina- 
tion was  made  for  a  series  of  seven  different  brightnesses,  from 
a  mere  glimmer  up  to  the  highest  capacity  of  the  apparatus. 

These  experiments  were  then  repeated  with  the  observer's  eye, 

*  A  paper  read  at  the  sixth  annual  convention  of  the  Illuminating  Engineering 
Society,  Niagara  Falls,  Out.,  September  16-19,  l9^2- 


cobb:   vision  as  influenced  by  surroundings 


293 


instead  of  being  in  the  darkness,  exposed  to  white  surface  illumi- 
nated as  nearly  uniformly  as  possible.  In  the  work  described 
this  was  accomplished  by  using  a  large  cube — 40  inches  (1.01  m.) 
side — with  its  edges  and  corners  filled.  The  result  of  this  con- 
struction was  a  hollow  figure  of  twenty-six  sides  each  tangent 
to  a  sphere  inscribed  in  the  original  cube.  The  interior  of  this 
was  painted  white  and  lit  by  a  100-watt  lamp  through  a  milk- 
glass  which  formed  one  of  the  oblique  sides.    The  observer  placed 


0.01  0.1  1.0  10.  42      100 

Brlqhtness-  eandlerpower  per  square  meter 

Fig.  1.— Variations  in  visual  angle  (V,)  and  brightness-difference-perception  (1^,  M])  with 
absolute  brightness  of  test-object,  remainder  of  field  is  absolutely  dark.  And  same 
(Vs,  I<s,  Ms)  surrounding  field  at  a  brightness  of  42  c-p.  per  square  meter.  Abscissae 
(brightness)  plotted  logarithmically,  figures  giving  actual  values.  Ordinates  in  the 
case  ot  visual  acuity  are  plotted  as  visual  angle  in  minutes,  subtended  by  centers  of 
adjacent  dark  and  bright  lines  in  tesl-object.  In  the  case  of  brightness-difference 
curves  the  ordinates  are  actual  differences  per  cent. 

his  face  in  an  opening  in  one  side  and  through  an  opening  directly 
opposite  him  he  looked  at  the  test-field.  The  milk-glass  through 
which  the  interior  was  illuminated  formed  part  of  the  oblique 
surface  over  his  head  and  outside  of  his 'visual  field,  so  that  he 
saw  nothing  but  the  white,  illuminated  inner  surface  of  the  fig- 
ure and  the  test-field.  With  the  eye  under  these  conditions  the 
experiments  were  repeated  with  known  brightnesses  of  the  test- 


294  TRANSACTIONS   I.  E.  S. — PART   II 

field,  this  latter  being  kept  wholly  independent  of  the  illumination 
of  the  interior  of  the  cube.  The  complete  results  are  given  in 
Fig.  i. 

Any  one  who  has  ever  attempted  to  determine  the  limiting 
value  of  a  stimulus  will  know  that  the  exact  point  of  division 
between  noticeability  and  the  reverse  is  to  be  arrived  at  only  as 
the  properly  determined  mean  of  a  large  number  of  judgments. 
At  or  near  the  critical  point  it  is  very  difficult  to  be  sure  whether 
one  sees  lines  or  not;  or  whether  there  is  a  difference  in  bright- 
ness or  not,  and  if  so  what  its  direction  is.  The  observer's  verdict 
is  in  any  one  case  obviously  to  a  certain  extent  at  the  mercy  of 
numerous  minor  influences,  and  of  the  state  of  his  own  mind. 
Further,  if  it  were  left  to  the  experimenter's  ex  tempore  judg- 
ment just  what  stimuli  to  present  to  the  observer  the  result  would 
obviously  depend  to  some  extent  on  the  experimenter's  preposses- 
sion in  that  respect.  He  would,  unconsciously  of  course,  be 
influenced  to  present  stimuli  at  a  time  when  he  anticipated  an 
answer  from  the  observer  that  coincided  with  his  own  anticipa- 
tions. 

For  these  reasons,  psycho-physicists  have  devised  special  meth- 
ods for  such  estimations.  A  detailed  discussion  of  psycho-physi- 
cal methods  would  be  out  of  place  here,  but  two  cardinal  points 
may  be  mentioned  with  a  brief  account  of  their  application  to 
the  present  work. 

(i)  After  determining,  by  a  few  preliminary  experiments, 
approximately  how  the  results  will  come  out,  the  method  of 
experimentation  is  planned  so  as  to  be  as  free  as  possible  from 
any  arbitrary  choice  on  the  part  of  either  the  experimenter  or  the 
observer.  The  whole  procedure  is  cut  and  dried  and  no  detail 
left  to  judgment  in  the  course  of  the  work  if  it  can  possibly  be 
planned  beforehand. 

(2)  When  unavoidable  circumstances  (such  as  whether  lines 
are  shown  to  the  observer  the  first  or  last  half,  of  an  experi- 
mental session)  would  probably  influence  the  result,  the  indi- 
vidual series  of  observations  are  so  grouped  that  one  half  will  be 
influenced  one  way,  the  other  half  the  contrary  way,  and  the  dis- 
tortion will  hence  average  out.    The  work  has  to  be  so  planned, 


cobb:   vision  as  influenced  by  surroundings        295 

of  course,  with  respect  to  every  circumstance  that  may  alter  the 
result. 

(1)  The  single  series  of  observations  consisted  of  sixteen 
exposures  or  stimuli.  In  the  case  of  lines,  for  example,  the 
finest  shown  in  any  one  series  were  distinctly  too  small  to  be  seen 
under  the  test-conditions,  the  coarsest  distinctly  visible,  and  the 
fifteen  steps  between  represented  a  series  of  equal  intervals  from 
the  one  to  the  other.  These  sixteen  stimuli  were  shown  to  the 
observer  in  haphazard  order,  previously  determined  by  lot,  and 
the  observer's  judgment  rendered  as  positive,  negative,  or  doubt- 
ful, as  to  visibility  of  the  lines,  and  recorded. 

Similarly,  in  the  case  of  brightness  difference  of  the  two  halves 
of  the  field,  the  two  extremes  of  each  series  represented  distinct 
difference  in  either  direction  (right  half  brighter  or  darker  than 
left)  with  equal  intervals  between. 

(2)  Each  session  consisted  of  four  series  for  each  of  the  two 
observers,  the  other  being  experimenter  at  the  time.  The  four 
series  were :  two  brightness-difference  series,  one  judged  as  to 
the  right  side,  one  as  to  the  left,  and  two  series  of  visual  acuity 
judgments.  It  is  in  the  making  up  of  a  session  such  as  this  that 
factors  enter  which  might  have  a  very  serious  influence  on  the 
result,  i.  e.,  (a)  whether  observer  A  came  before  or  after  B, 
{b)  whether  he  was  shown  lines  before  brightness-difference  or 
after,  and  (c)  whether  he  judged  left  or  right  side  first  (in  the 
case  of  brightness-difference  only).  The  entire  set  was  there- 
fore carefully  and  systematically  planned  so  that  A  and  B  came 
first  each  in  an  equal  number  of  series,  lines  and  brightness- 
difference  sharing  precedence  in  a  similar  way,  and  in  the  case 
of  brightness-difference,  judgment  on  the  right  and  left  side  each 
came  first  in  just  one  half  of  the  series. 

The  various  intensities  of  the  test-field  also  had  to  be  dis- 
tributed each  in  a  fairly  representative  way  over  the  whole  period 
of  the  complete  set,  in  order  that  any  changes  in  the  eye  that  take 
place  over  the  entire  period  of  weeks  or  -months  may  take  effect 
alike  on  the  observations  at  all  the  intensities. 

The  interpretation  of  the  results  given  in  the  curves  is  as 
follows : 

Visual  acuity  (V)  is  plotted  as  the  smallest  visual  angle  in 
7 


296 


TRANSACTIONS    I.  E-  S. — PART    II 


minutes  (from  the  center  of  any  black  line  to  the  center  of  the 
adjacent  bright  line)  under  which  the  lines  can  be  distinguished. 

Brightness  difference  is  plotted  as  (L)  the  limen,  that  is  the 
difference  per  cent,  which,  by  inference  from  the  results  actually 
obtained  would  be  correctly  recognized  in  just  one  half  of  the 
cases  in  which  it  was  presented  to  the  observer,  and  (M)  the 
average  difference  at  which  in  each  series  the  first  deviation  from 
"correct"  judgment  occurs. 

The  results  given  are  the  means  of  all  the  observations  of  the 
two  observers  in  each  case.  In  spite  of  the  precautions  outlined, 
the  curves  in  the  figure  show  certain  apparently  erratic  varia- 
tions.   Whether  these  are  essential  or  accidental  does  not  appear. 


s 

> 

f\s 

,'-° 

/ 
/ 
• 

as 

ac 

1              0 

1               1. 

3                       IC 

100 

Briqhtness- candle-power  per  square  meter 

Fig.  2. — Visual  acuity  curves  derived  from  the  same  values  as  V!  and  V2  in  fig.  1. 


The  only  way  to  determine  that  point  would  be  to  multiply  the 
number  of  experiments.  The  results  given  do,  however,  show 
clearly  enough  certain  features. 

The  rapid  increase  in  the  least  noticeable  difference  at  very 
low  intensities  as  shown  years  ago  by  Koenig,  comes  out  clearly 
in  the  curves  Mt  and  Lx  (dark  background)  as  also  the  low  value 
for  visual  acuity  (large  visual  angle)  at  low  intensities  (V^. 

When  the  bright  background  is  used  the  lower  limit  of  vision 
is  evidently  pushed  up  to  a  much  higher  intensity  as  definitely 
shown  in  M2  and  L2,  and  intimated  in  V2  by  its  altered  trend. 


cobb:    vision  as  influenced  by  surroundings         297 

By  far  the  most  striking  point  in  the  whole  work  is,  to  the 
writer's  mind,  however,  the  fact  that  by  all  three  criteria  used, 
vision  at  the  highest  intensity  of  test  object  shows  a  distinct 
improvement  in  the  presence  of  the  bright  visual  field,  i.  e.,  vision 
is  actually  improved  by  filling  the  visual  field  with  surface  almost 
as  bright  as  the  test-object.  The  curve  showing  the  visual  angle 
translated  into  visual  acuity  (Fig.  2)  brings  this  out  more  clearly. 

Under  certain  circumstances  then,  the  eye  can  see  more  clearly 
when  a  large  amount  of  light  falls  into  it  than  when  this  is  cut 
off,  or  in  other  words,  the  eye  can  see  equally  clearly  under  those 
circumstances  with  a  smaller  amount  of  light  on  the  test-object. 
That  is,  what  has  been  called  the  "efficiency  of  the  eye"  is  in 
some  cases  greater  with  a  relatively  large  amount  of  light  coming 
to  the  eye  not  from  the  test-object. 

DISCUSSION. 

Mr.  L.  B.  Marks  :  It  has  been  found  in  practise  that  an 
extremely  concentrated  lighting  field  with  a  dark  background 
produces  visual  fatigue.  Numerous  cases  of  this  kind  have  come 
to  my  notice  in  connection  with  factory  and  other  lighting  work. 

In  considering  wall  color  and  brightness  of  objects  within  the 
field  of  view,  one  must  take  into  account  the  actual  intensity  of 
illumination  and  the  ratio  of  intensities.  In  Dr.  Cobb's  analysis, 
for  example,  if  we  take  very  low  intensities  it  might  easily  follow 
that  a  darker  wall  would  be  preferable  to  a  lighter  one,  while  if 
we  take  the  higher  intensities,  the  lighter  wall  would  be  preferable. 

Mr.  J.  R.  Cravath  :  The  point  which  interests  me  especially 
in  Dr.  Cobb's  paper  is  that  of  the  value  of  illumination  on  the 
background ;  namely,  42  candle-power  per  square  meter,  is  a  value 
which  is  frequently  attained  in  practise  on  light  colored  walls. 
We  would  have  good  reason  to  suppose,  with  the  information 
which  the  paper  gives,  that  we  would  have  higher  visual  acuity 
under  practical  working  conditions  with  such  rather  highly  illumi- 
nated walls.  Dr.  Cobb  does  not  attempt  to  explain  why  this  is 
so.  It  has  been  suggested  that  the  diameter  of  the  pupil  is  less 
where  we  have  bright  surroundings  and  therefore  we  are  able  to 
focus  more  clearly,  just  as  we  get  a  clearer  image  in  the  camera 
with  a  small  aperture. 


298  TRANSACTIONS   I.  E.  S. — PART   II 

Dr.  H.  E.  Ives:  I  have  a  little  fact  to  bring  forward  in  line 
with  Dr.  Cobb's  research,  in  connection  with  the  photometry  of 
lights  of  different  colors.  You  may  have  noticed  that  in  my 
paper  this  morning  I  recommended  that  the  photometric  field  be 
surrounded  by  a  larger  field  of  equal  brightness.  I  did  not  go 
into  details  in  my  abstract  this  morning  in  regard  to  that,  but  I 
might  enlarge  upon  it  now. 

In  comparing  lights  of  different  colors  by  the  equality  of  bright- 
ness method,  I  found  by  experiment  that  the  presence  of  bright 
surroundings  to  the  photometric  field  resulted  in  the  range  of 
settings  being  approximately  divided  by  two ;  in  other  words,  the 
sensibility  was  enormously  increased. 

As  regards  the  question  of  fatigue,  it  has  been  my  experience 
that  the  eye  becomes  very  much  fatigued  looking  down  the  black 
tube  of  the  ordinary  optical  instrument  at  a  bright  field ;  on  the 
other  hand,  using  a  bright  tube,  one  could  work  very  much  longer. 
There  was  really  a  tremendous  difference.  In  the  paper  this 
morning  this  observation  was  incorporated  and  the  reason  given 
is  that  it  makes  for  the  comfort  of  the  observer.  I  am  giving 
this  for  what  it  is  worth.  My  own  interpretation  of  it  is  that 
bright  walls  are  not  quite  as  black  as  they  have  been  painted. 

I  do  not  think  any  difference  in  the  diameter  of  the  pupil  would 
explain  the  differences  found  in  the  discrimination  of  shade  differ- 
ences, which  show  the  same  characteristics  as  the  visual  acuity 
tests. 

Mr.  H.  P.  Gage:  I  would  just  like  to  say  a  word  about  the 
last  part  of  the  paper,  where  Dr.  Cobb  shows  that  it  is  possible 
to  detect  differences  in  brightness  and  fineness  of  detail  with  a 
brighter  surrounding  field  than  with  a  dark  surrounding  field. 
I  think  this  is  due  to  the  fact  that  the  eye  is  able  to  focus  accu- 
rately on  the  test  object,  whereas  with  a  dark  surrounding  field, 
the  eye  is  out  of  focus  and  the  object  is  not  seen  clearly. 

Another  point,  a  little  outside  of  this,  was  brought  out  by  the 
use  of  the  acuity  method,  showing  that  a  monochromatic  light 
will  give  a  higher  acuity  than  white  light.  This  has  been  pretty 
well  recognized  as  due  to  the  chromatic  aberration  of  the  eye. 
It  would  be  very  interesting  to  try  some  of  these  tests,  using  a 
glass  in  front  of  the  eye  designed  to  make  the  eye  an  achromatic 
instrument.    The  normal  eye  at  rest  will  be  found  to  be  in  focus 


VISION  AS  INFLUENCED  BY  SURROUNDINGS  299 

for  the  red  end  of  the  spectrum,  if  the  eye  is  correctly  focused 
for  a  red  light  and  it  will  be  found  that  for  blue  or  green  light, 
the  eye  will  be  short-sighted.  This  has  an  important  bearing  on 
signalling. 

The  other  night  I  had  a  negative  lens  and  I  took  occasion  to 
examine  the  lights  from  the  rear  of  a  train  (red,  green  and 
yellow).  It  was  interesting  to  note  the  greater  range  at  which 
the  red  light  could  be  seen  under  the  ordinary  conditions  of  the 
eye  at  rest.  With  the  negative  lens,  which  renders  the  eye  far- 
sighted,  the  green  light,  which  appeared  as  a  bright  point  with 
rays  around  it,  was  brought  down  to  a  definite  round  spot  and 
could  be  seen  at  a  much  greater  distance  than  without  the  glass. 
This  indicates  the  reason  why  different  people,  not  color  blind, 
will  get  different  appearances  with  the  same  colored  signals. 

In  signal  work,  not  only  the  question  of  color  sensitiveness  and 
color  blindness  enters  into  consideration,  but  also  the  dioptric  con- 
dition of  the  eye  whether  near  or  far-sighted  and,  if  normal,  for 
which  particular  wave-length  the  eye  is  in  focus  when  at  rest. 

Mr.  M.  Luckiesh  :  The  data  given  show  that  under  the  con- 
ditions of  the  test,  better  results  are  obtained  amid  bright  sur- 
roundings when  the  brightness  of  the  object  is  10  to  100  candle- 
power  per  square  meter.  The  bright  surroundings  were  not  any 
brighter  than  those  very  often  found  in  actual  practise. 

In  regard  to  fatigue :  the  periods  of  the  test  were  about  45 
minutes  long  for  the  observer.  This  is  not  considered  by  psy- 
chologists to  be  long  enough  to  induce  undue  fatigue.  In  fact 
Dr.  Cobb's  data  do  not  show  any  effects  of  fatigue  when  the  data 
of  the  first  and  last  halves  of  the  period  are  compared. 

The  question  of  better  acuity  in  monochromatic  light  was  quite 
thoroughly  thrashed  out  at  the  convention  of  the  Society  last 
year. 

Dr.  P.  W.  Cobb  (in  reply)  :  Referring  to  Fig.  2 — at  the  level 
where  visual  acuity  is  1.5  the  bright  surroundings  evidently  from 
the  (dotted)  curve  gave  the  eye  an  advantage,  since  vision  main- 
tained this  value  with  a  less  bright  test  object  than  was  necessary 
in  dark  surroundings.  That  is,  the  "efficiency  of  the  eye"  was 
greater  than  unity — in  fact  (bearing  in  mind  that  we  are  using  a 
logarithmic  abscissa-scale)  the  "efficiency  of  the  eye"  estimated 


300  TRANSACTIONS    I.  E.  S. — PART    II 

from  the  curves  would  be  here  about  2.00.  At  5  candle-power 
per  square  meter  (about  one-eighth  of  the  surrounding  bright- 
ness) the  two  curves  cross — which  means  that  this  value  is  here 
unity,  and  below  this  point  it  is  less  than  unity.  It  could  evidently 
be  made  almost  indefinitely  small  by  selecting  a  low  enough  test- 
object  brightness  to  start  with.  The  curves  show  it  to  be  about 
0.15  at  the  lowest  point  given  (visual  acuity  a  trifle  over  1.0). 

I  would  add  that  the  visual  acuity  values  came  out  much  more 
consistently  and  definitely  than  the  brightness-difference  values 
and  are  hence  much  more  trustworthy.  However,  as  the  curves 
show,  the  change  in  conditions  works  a  relative  change  in  results 
in  every  case  of  the  same  direction  as  that  seen  in  the  case  of 
visual  acuity. 

I  admit  that  a  more  thorough  investigation  of  the  matter  in 
the  region  of  higher  brightness  is  desirable  as  far  as  the  M  and  L, 
curves  are  concerned.  That  is  easier  seen  than  foreseen.  Each 
point  in  the  figure  indicated  by  a  small  circle  is  the  result  of  512 
(in  a  few  cases  only  448)  separate  judgments,  each  pronounced 
after  3  seconds  observation  of  the  test  object.  The  entire  work 
of  which  these  are  the  final  results  extended  over  the  larger  part 
of  a  year. 

In  reply  to  Dr.  Gage,  I  would  say  that  the  surrounding  surface 
was  1  meter  distant,  while  the  test-object  itself  was  2  meters 
from  the  eye.  In  view  of  this,  accommodation  for  the  test-object 
should  be  more  disturbed  when  this  latter  was  lit  up.  There  was 
no  evidence  whatever  of  any  difficulty  of  this  kind.  The  blank 
field  was  always  present  to  the  eye  before  the  exposure,  having 
borders  identical  with  those  of  the  test-field  which  served  as  a 
guide  to  the  eye  in  accommodation. 

The  observation  cited  by  another  speaker  is  interesting.  The 
conditions  in  the  case  he  mentions  (light  from  a  source  directly 
upon  the  eye)  are  not  the  same  as  an  equal  number  of  meter- 
candles  coming  from  an  extended  and  uniformly  bright  surface 
to  the  eye.  As  a  matter  of  fact,  working  with  the  brightest  test- 
object  on  a  dark  field  produced  considerable  discomfort  and 
vaguely  uncertain  feelings  referable  to  the  eyes,  which  did  not 
appear  in  the  presence  of  the  bright  surroundings. 

It  is  exactly  such  ideas  as  those  that  Mr.  Marks  has  just 
expressed  that  have  made  this  work  appear  to  me  to  be  worth 


VISION   AS   INFLUENCED  BY  SURROUNDINGS  301 

while.  Nature,  rarely  if  ever,  treats  man  to  an  extreme  local  type 
of  illumination.  We  must  bear  in  mind  that  sunlight  is  about 
one  thousand  times  as  intense  as  ample  artificial  illumination ;  so 
I  do  not  feel  that  the  brightness  chosen  for  the  background  in 
this  work  is  at  all  abnormal  to  the  eye,  although  it  may  not  fall 
within  the  limits  of  engineering  practise. 

Replying  to  Mr.  Cravath :  I  have  preferred  in  this  paper  to 
give  the  results  without  going  into  the  theory  of  the  matter.  A 
smaller  pupil  does  not  necessarily  mean  more  distinct  vision 
except  where  the  refraction  of  the  eye  is  imperfect.  In  my  own 
eye  (moderate  astigmatism,  corrected)  I  found  by  using  a  set 
of  artificial  pupils  that  visual  acuity  grew  markedly  less  with 
decrease  in  pupillary  size  below  3  mm.,  even  when  the  brightness 
of  the  test-object  was  raised  to  compensate  exactly  the  decrease 
in  pupillary  area.  And  further,  the  M  and  L  curves,  Fig.  1, 
show  changes  in  the  same  direction  that  visual  acuity  does,  and 
it  is  fair  to  suppose  that  exactness  of  the  retinal  image  plays  at 
most  a  much  less  significant  part  in  the  estimation  of  brightness 
differences  in  visually  gross  areas.  I  do  not  wish  to  be  under- 
stood, however,  as  saying  that  the  pupillary  diameter  has  been 
eliminated  as  a  factor.  I  do  feel  that  other  things  are  necessary 
for  a  full  explanation. 

There  is  in  progress  at  the  laboratory  with  which  I  am  con- 
nected further  work  of  similar  character  to  this,  using  a  bright- 
ness of  3  candles  per  square  meter  for  the  surroundings  instead 
of  42,  which  it  is  hoped  will  make  the  entire  question  more  clear. 


302  TRANSACTIONS    I.  E.  S. — PART    II 

A    PRACTICAL    SOLUTION    OF    THE    PROBLEM    OF 
HETEROCHROMATIC  PHOTOMETRY 


BY   PROF.    CH.   FABRY. 


Synopsis:  In  order  to  reduce  all  practical  photometry  to  that  of  lights 
of  the  same  color,  secondary  standards  of  various  colors  or  colored 
absorbing  media  are  necessary.  These  should  be  calibrated  in  the 
standardizing  laboratory,  by  the  most  refined  methods  of  heterochromatic 
photometry.  The  calibration  of  an  infinity  of  colored  standards,  or  of 
every  individual  colored  glass,  is  not  practical.  Herewith  are  described 
absorbing  liquids,  definitely  specifiable,  which  may  be  used  in  varying 
thicknesses  and  proportions  to  make  the  light  of  a  given  standard  like 
that  of  any  other  illuminant.  A  yellow  and  a  blue  solution  have  been 
found  which  suffice  to  match  with  a  Carcel,  all  the  ordinary  illuminants; 
a  purple  solution  is  suggested  for  use  where  these  are  not  sufficient,  the 
three  absorptions  giving,  by  the  three-color  principle,  all  the  tints  as 
represented  in  a  color  triangle.  The  method  of  use  of  these  screens  is 
outlined,  the  possible  methods  of  calibration  are  described  and  some 
experimental  results  are  tabulated. 

In  spite  of  many  good  works,  the  problem  of  heterochromatic 
photometry  seems  not  to  be  satisfactorily  solved  for  practical  use. 
In  the  comparison  of  two  lights  of  very  different  colors,  as,  for 
instance,  an  incandescent  carbon  lamp,  used  as  a  standard,  with 
daylight  or  the  mercury-vapor  arc,  there  remains  a  large  amount 
of  uncertainty  and  arbitrariness.  Several  very  good  papers, 
published  in  these  Transactions,  especially  by  Dr.  Ives,  have 
forwarded  the  question.  Probably  the  definitive  solution  will 
be  conformable  to  the  suggestion  of  Dr.  Ives1  namely :  Measure 
every  simple  radiation  in  quantity,  by  the  energy  conveyed  by 
it  in  one  second,  an  operation  entirely  independent  of  the  prop- 
erties of  the  human  eye;  then,  quote  every  one  at  its  proper 
value  for  the  special  application  to  lighting,  this  value  being  zero 
for  all  infra-red  or  ultra-violet  radiation.  Thus,  for  computing 
the  value  of  some  ore  containing  metals  of  different  prices  (gold. 
silver,  copper,  etc.)  one  will  determine  the  weight  of  each  metal, 
and  each  one  will  be  quoted  at  its  proper  price,  sand  and  stone 
being  quoted  at  zero  for  metallurgic  purposes. 

For  this  definitive  solution  we  are  not  yet  ready,  at  least  for 

1  Transactions  of  the  Illuminating  Engineering  Society  April,  1911,  page  258,  and 
Astrophysical  Journal.  November,  1912,  page  322. 


Fabry:   problem  of  heterochromatic  photometry     303 

everyday  use.  Meanwhile,  we  are  obliged  to  choose  some  pro- 
visional solution,  and  a  French  proverb  says  that  nothing  lasts 
longer  than  a  makeshift.  It  is  moreover  possible  that  a  pro- 
visional solution  could  be  turned  into  a  definitive  one  when  it 
will  be  possible  to  properly  express  the  experimental  results. 

In  actual  practise,  we  cannot  think  of  making  a  complete  analy- 
sis, qualitative  and  quantitative  of  every  light  to  be  measured. 
So  long  as  the  eye  will  be  used  in  photometric  measurements,  the 
equalization  of  two  illuminations  on  the  photometric  screen  is 
all  that  can  be  asked  for  the  ordinary  observer,  and  this  equaliza- 
tion can  be  accurately  made  in  only  one  case,  i.  e.,  when  the  lights 
are  of  the  same  color.  We  are  thus  led  to  eliminate  from  prac- 
tise any  heterochromatic  measurement,  as  was  excellently  ex- 
pressed by  Dr.  Ives  :2  "I  think  it  extremely  important — it  is 
essential — that  in  ordinary  photometry  there  should  never  be 
made  a  comparison  of  different  colors.  All  practical  photometry 
should  be  reduced  to  photometry  of  the  same  color.  Conse- 
quently, the  question  of  which  photometer  is  to  be  used  for 
comparing  lights  of  different  colors,  becomes  a  question  for  the 
standardizing  laboratory,  the  Bureau  of  Standards,  the  Reich- 
sanstalt  or  the  National  Physical  Laboratory." 

What  have  we  to  do  in  order  to  reach  this  end,  viz.,  to  have 
only  comparisons  between  lights  of  the  same  color?  We  cannot 
hope  to  find  as  many  different  standards  as  there  are  different 
tints,  i.  e.,  an  infinity :  the  difficulties  encountered  in  finding  one 
good  standard  do  not  encourage  one  to  search  so  large  a  number. 

With  a  single  photometric  standard,  the  only  chance  of  suc- 
cess is  to  modify  its  tint  by  interposition  of  some  apparatus  in 
the  beam  of  light,  in  order  to  give  it  the  desired  color.  This 
change  of  color  will  at  the  same  time  modify  the  intensity  in  a 
ratio  which  will  be  measured  once  for  all  in  every  case ;  this 
measurement,  unavoidably  involving  heterochromatic  measure- 
ment, will  be  made  by  the  standardizing  bureau. 

There  are  many  means  at  the  hand  of,  physicists  by  which  the 
tint  of  a  complex  radiation,  i.  e.,  the  proportion  of  the  different 
simple    radiations    in    the    light    under    consideration,    may    be 

2  Transactions  of  the  Illuminating  Engineering  Society,  November,  1910,  page  727. 


304  TRANSACTIONS    I.  E.  S. — PART    II 

changed  at  will.3  From  a  practical  standpoint,  the  simplest 
means  is  the  interposition  of  some  absorbing  medium.  The  use 
of  such  absorbing  substances  (colored  glasses)  has  been  sug- 
gested and  actually  tried.4  The  procedure  is  as  follows :  We 
have,  for  instance,  to  compare  an  electric  arc  with  an  incandes- 
cent lamp  used  as  a  standard ;  when  the  two  parts  of  the  photo- 
metric screen  are  illuminated  by  the  two  sources,  they  appear 
very  different  in  color,  and  the  equalization  of  brightness  is 
largely  a  matter  of  judgment.  On  the  incandescent  lamp  a  blue 
glass  properly  chosen  is  superposed,  which  absorbs  the  excess 
of  red  and  yellow  light  contained  in  the  radiation  of  this  lamp, 
and  gives  it  the  same  color  as  that  of  the  electric  arc.  The 
equalization  of  illumination  on  the  photometric  screen  will  then 
become  easy  and  precise.  By  this  procedure,  we  are  able  to  com- 
pute the  intensity  of  the  arc  lamp  if  we  know  how  many  candles 
the  incandescent  lamp  gives  through  the  blue  glass.  As  we 
know  the  intensity  of  the  lamp  without  the  absorbing  glass,  we 
need  only  to  have  the  ratio  between  incident  light  and  trans- 
mitted light  through  the  glass,  i.  e.,  the  opacity  of  this  glass  for 
the  total  radiation  of  this  lamp.  The  opacity  can  be  measured 
once  for  all,  but  this  value  is  right  only  for  the  glass  employed 
and  for  a  lamp  giving  the  same  kind  of  radiation  as  the  lamp 
used  for  the  measure.  If  a  higher  voltage  is  applied  to  the  lamp, 
the  proportion  of  blue  radiations  freely  transmitted  through  the 
glass  will  be  greater,  and  the  opacity  will  be  less.  The  measure 
of  these  opacities  for  fixed  light  would  be  the  problem  of  the 
standardizing  Bureau,  and  as  it  is  probably  impossible  to  repro- 
duce glasses  exactly,  it  would  be  necessary  to  make  the  measure- 
ment on  every  separate  sample  of  glass,  of  which  a  great  many 
would  be  necessary  to  match  the  color  of  the  different  sources 
now  employed  in  practise. 

An  important  step  would  be  reached  if  it  were  possible  to  use 

3  All  these  means  work  by  inequal  weakening  of  the  different  simple  radiations; 
they  are  incapable  of  adding  simple  radiations  lacking  in  the  considered  light.  If  this 
was  monochromatic,  any  one  of  these  means  could  not  change  its  composition.  We  con- 
sider, in  fact,  as  standard  source,  such  a  source  as  giving  forth  a  mixture  of  every  simple 
luminous  radiation.  In  such  a  case,  it  is  conceivable  that,  by  proper  weakening  of  the 
different  simple  radiations,  it  would  be  possible  to  get  every  possible  mixture  of  simple 
radiations,  i.  e.,  every  possible  light. 

4  Ives,  Transactions  of  the  Illuminating  Engineering  Society,  November.  1910,  page 
728;  Cady,  October,  1912,  page  385. 


FABRY  :     PROBLEM   OF    HETEROCHROMATIC    PHOTOMETRY      305 

definite  absorbing  media,  reproducible  by  every  observer;  the 
opacities  of  different  thicknesses  might  be  measured  once  for  all, 
and  every  observer  would  be  able  to  use  these  values,  without 
purchasing  a  definite  object  measured  at  the  bureau.  The  light 
used  for  these  measures  of  opacity  should  be  of  definite  com- 
position, but  it  need  not  be  the  light  of  a  standard  of  intensity ; 
it  is  possible,  and  perhaps  better,  to  separate  the  two  duties,  in 
order  to  choose  the  best  standard  for  each  purpose ;  the  standard 
of  intensity  ought  to  have  a  fixed  intensity,  and  a  very  small 
change  of  tint  from  one  standard  to  another  is  of  no  conse- 
quence ;  the  standard  of  color  must  give  a  radiation  of  constant 
composition,  but  a  change  of  intensity  from  day  to  day  is  abso- 
lutely unimportant  because  it  will  every  time  be  compared  with 
the  standard  of  intensity. 

These  thoughts  came  in  my  mind  ten  years  ago,  and  I  made  a 
long  series  of  experiments  to  put  them  into  practise.  Some 
physicists  to  whom  I  showed  my  results  raised  many  objections, 
from  which  I  concluded  that  the  theory  of  colors  was  not  yet 
sufficiently  understood  to  permit  the  general  use  of  my  method ; 
I  published  only  a  short  article,3  which  was  entirely  unnoticed. 
The  situation  is  now  changed,6  and  I  think  it  is  now  worth  while 
to  publish  a  more  complete  account  of  my  experiments,  in  the 
hope  of  contributing  to  the  improvement  of  the  methods  of 
heterochromatic  photometry,  which  are  still  in  a  somewhat  chaotic 
condition. 

I  suppose  the  question  of  the  standard  of  intensity  to  be  solved. 
In  the  practise  of  to-day,  the  standard  is  not  a  flame,  but  an 
incandescent  carbon  lamp  which,  at  fixed  voltage,  gives  a  definite 
number  of  candles. 

Furthermore,  I  chose  a  source  giving  light  of  definite  tint,  with- 
out respect  to  intensity,  which  need  only  remain  constant  during 
the  short  time  of  the  measure.     I  have  chosen  the  old  Carcel 

6  Comptes-Rendus  de  V  Aeademie  des  Sciences,  9  Novembre,  1903,  (Sur  une  solution 
pratique  du  probleme  de  la  photomfitrie  heterochrome). 

*  If,  today,  the  technical  men  are  better  prepared  to  understand  the  problems  of 
heterochromatic  photometry,  that  is  in  a  large  degree  owing  to  the  works  published  by 
the  Illuminating  Engineering  Society  ;  it  is  for  that  reason  that  I  have  wished  my  paper 
to  be  published  in  these  Transactions,  the  readers  of  which  are  the  best  prepared  to 
grasp  the  problem  here  dealt  with. 


306  TRANSACTIONS    I.  E.  S. — PART    II 

lamp,7  which  is  probably  good  enough  for  this  very  simple  ser- 
vice; perhaps  a  carbon  incandescent  lamp  at  a  fixed  number  of 
watts  per  candle  would  be  more  convenient.  The  light  of  the 
Carcel  is  very  similar  to  that  of  the  carbon  lamp  ordinarily  used 
as  standard  of  intensity,  but  is  very  different  from  many  other 
lights,  especially  from  the  light  of  the  sun. 

I  have  sought  such  colored  substances  as,  interposed  in  the 
light  of  the  Carcel,  will  give  it  the  color  desired  in  every  case. 
To  have  reproducible  absorbing  media,  whose  opacity  can  be 
measured  once  for  all,  one  is  lead  to  use  liquids  of  definite  com- 
position and  proper  thicknesses.  It  is  desirable  to  be  able  to  get 
every  possible  tint,  in  order  to  match  the  color  of  the  standard 
with  that  of  each  light.  A  single  liquid  is  not  sufficient  to  arrive 
at  this  end;  if  only  one  liquid  is  used,  only  one  variable  is  at 
one's  disposal,  viz.,  the  thickness  of  this  substance  or,  what  is 


the  same,  the  concentration  of  the  absorbing  dye  in  mixture  with 
an  uncolored  liquid  as  water.  A  single  set  of  colors  can  be  thus 
obtained ;  if  every  tint  is  represented  by  a  point  in  the  triangular 
diagram  of  Maxwell-Koenig,s  the  different  tints  produced  in  vary- 
ing the  thickness  will  be  represented  by  the  different  points  on 
the  curve  A  (Fig.  i)  beginning  at  the  point  M  which  represents 
the  tint  of  the  Carcel  without  the  absorbing  medium  (thickness 
zero).  Only  the  colors  represented  by  some  point  on  the  curve  A 
can  be  matched  by  the  use  of  the  Carcel  with  the  absorbing 
medium  under  consideration. 

Another  liquid  of  different  color  will  give  another  set  of  tints 

7  I  did  not  use  the  Hefner  lamp,  on  account  of  its  too  red  color,  and  because  it  is  too 
weak  when  further  weakened  by  absorption.  The  light  of  the  Carcel  is  so  similar  to  that 
of  the  carbon  incandescent  lamp  at  normal  voltage  that  the  values  given  below  would 
probably  be  applicable  without  change  to  this  latter  source. 

B  For  the  definition  of  this  diagram,  see  :  Ives,  Transactions  of  the  Illuminating 
Engineering  Society;  April,  1910,  page  205,  and  April,  1911,  page  266. 


FABRY  :     PROBLEM   OF    HETEROCHROMATIC    PHOTOMETRY      307 

represented  on  the  diagram  by  the  different  points  on  another 
curve  B  beginning  at  the  same  point  M. 

If  two  absorbing  cells  are  used,  one  of  the  first  liquid  and  one 
of  the  second,  we  have  at  will  the  values  of  two  variables,  the 
thicknesses  of  the  two  cells ;  it  will  be  possible  to  obtain  all  the 
tints  represented  in  a  certain  part  of  the  diagram.  We  cannot 
cover  all  the  surface  of  the  diagram,  but,  at  least,  a  finite  surface 
of  it.  If  the  two  liquids  are  properly  chosen,  we  can  hope  to 
match  exactly  the  light  of  many  of  our  ordinary  sources. 

The  two  liquids  chosen  are  one  of  blue  color,  with  absorption 
increasing  from  violet  to  red,  and  one  yellow  with  absorption 
varying  in  the  inverse  direction.    The  aim  to  have  liquids  definite 


1  i 

1  1 

1  1 

Fig.  2. 

and  reproducible  leads  us  to  set  aside  all  the  anilin  dyes,  and  to 
seek  only  among  the  mineral  substances.  The  two  liquids  chosen 
were : 

A.  Crystallized  copper  sulphate 1  gram 

Commercial  ammonia  (density  0.92) 100  cubic  centimeters 

Water,  quantity  sufficient  to  make  1   liter 

B.  Potassium  iodide  3  grams 

Iodine    1  gram 

Water,  quantity  sufficient  to  make  1  liter 

The  thicknesses  of  the  cells  containing  each  of  these  liquids  can 
be  varied  at  will ;  I  shall  denote  by  x  the  thickness  of  the  blue 
liquid  A  and  by  y  the  thickness  of  the  yellow  liquid  B,  the  two 
expressed  in  millimeters.  The  thickness  can  be  varied  at  will  if, 
for  each  liquid,  a  double  wedge  shaped  cell  is  used  (Fig.  2)  ; 
when  one  of  the  wedges  is  slid,  the  totar  thickness  of  the  liquid 
can  be  varied  continuously.  It  is  ordinarily  simpler  to  leave 
constant  the  thickness,  using  a  parallel  cell,  and  change  the  con- 
centration of  the  liquid ;  I  have  verified  the  fact  that  a  change  in 


308  TRANSACTIONS    I.  E.  S. — PART    II 

concentration  produces  the  same  effect  as  a  change  of  thickness 
(Beer's  law)  ;  x  and  y  are  then  the  products  of  the  thickness 
measured  in  millimeters  by  the  number  of  grams  per  liter  (copper 
•ulphate  or  iodine)  in  the  solution  used.  In  this  change  of  con- 
centration, one  must  leave  approximately  unchanged  the  propor- 
tion of  ammonia  contained  in  the  liquid  A,  and  consequently  add 
to  this  liquid  not  pure  water  but  water  containing  ammonia  in 
the  proportion  of  ioo  cubic  centimeters  to  the  liter.  For  the 
liquid  B,  pure  water  can  be  added.  I  ordinarily  use  cells  with 
parallel  faces  and  of  20  millimeters  thickness. 

With  these  two  absorbing  media  of  proper  thickness  and  con- 
centration, and  the  Carcel  lamp  used  as  standard  of  color,  it  will 
be  possible  to  match  the  color  of  almost  every  source  used  in 
practise.  With  a  single  cell  containing  the  liquid  A  (y  =  o)  we 
obtain,  with  increasing  thickness  (x  greater  and  greater)  light 
of  color  more  and  more  blue ;  we  can  match  in  this  way  the  color 
of  every  source  similar  to  the  black  body  at  higher  temperature 
than  the  Carcel,  including  the  solar  light.  The  tints  produced 
in  this  way  are  represented  in  the  diagram  (Fig.  1)  by  the  differ- 
ent points  on  the  curve  A.  With  a  single  cell  filled  with  the 
liquid  B,  by  increasing  the  thickness  or  concentration  of  this 
liquid  (x  =  o,  y  increasing),  we  obtain  lights  of  more  and  more 
yellow  and  finally  a  red  tint;  the  colors  so  obtained  are  repre- 
sented on  the  diagram  by  the  curve  B.  Using  the  two  cells  with 
every  possible  thickness  or  concentration  (x  and  y  varying  at 
will),  we  can  produce  light  represented  to  the  left  and  above  the 
curve  AMB,  viz.,  all  the  set  of  blue,  green,  yellow  or  red  tints, 
more  or  less  saturated.  We  cannot  get  the  tints  represented  to 
the  right  and  under  AMB,  i.  e.,  the  purple  tints,  in  which  are 
predominant  the  radiations  of  the  two  ends  of  the  spectrum  with 
weakening  of  the  middle  part ;  a  third  liquid,  used  with  one  or 
the  other  of  the  two  just  described,  would  be  necessary  in  order 
to  get  these  tints.  I  did  not  think  it  worth  while  to  seek  such  a 
liquid. 

For  every  source  of  light  we  have  to  search  for  the  values  of 
x  and  y  necessary  to  match  its  color.  The  following  table  gives 
some  values.9 

9  These  values  are  given  here  only  as  an  indication,  inasmuch  as  the  indicated  sources 
are  by  no  means  in  themselves  definite  in  color. 


FABRY  :     PROBLEM    OF    HETEROCHROMATIC   PHOTOMETRY      309 

x  y 

(Blue  (Yellow 

Source  of  light                                          solution)  solution) 

Ordinary  carbon  arc  (not  mineralized) 33  0.0 

Tungsten  lamp 10  0.0 

Acetylene  flame 12  0.0 

Incandescent  carbon  lamp  at  */j  of  its  normal 

voltage o  5.4 

Id.  at  :|  ,  of  its  normal  voltage 19  0.0 

Hefner  lamp o  1.5 

Auer  light 41  1.0 

Nernst  lamp 22  0.6 

Cooper-Hewitt  lamp 76  0.6 

Sun  at  noon  in  summer 54  0.0 

As  can  be  seen,  the  blue  solution  is,  in  most  cases,  the  most 
important,  and  indeed  in  many  instances,  must  be  used  alone  to 
produce  the  desired  tint.  It  is  the  case  for  all  sources  of  light 
whose  spectrum  is  similar  to  that  of  a  black  body  at  higher  tem- 
perature than  the  Carcel  (tungsten  lamp,  electric  arc,  acetylene, 
sun).  The  precision  with  which  these  sources  of  light  can  be 
matched  by  means  of  the  Carcel  and  copper  liquid  is  really  sur- 
prising. The  tint  of  the  sunlight,  so  greatly  different  from  that 
of  the  Carcel  as  to  render  almost  photometric  comparison  im- 
possible, is  practically  indistinguishable  from  the  color  of  the 
modified  Carcel.10  That  does  not  prove  that  the  energy  curves 
of  these  two  lights  are  entirely  identical,  but  the  energy  curve 
of  the  so  modified  Carcel  is  much  more  like  that  of  the  sun  than 
is  the  energy  curve  of  the  natural  Carcel,  as  is  to  be  seen  from 
Fig.  3  where  are  represented  the  energy  curves  of  the  Carcel  lamp, 
of  the  same  source  through  54  millimeters  of  copper  solution  and 
that  of  sunlight.  For  a  normal  eye  the  two  last  tints  are  identical, 
and  it  is  probably  so  for  every  eye  not  entirely  abnormal.  The 
white  stars  (like  Vega)  give  a  light  still  more  blue  than  that  of 
the  sun ;  they  could  probably  be  matched  with  the  Carcel  light 
by  using  a  somewhat  greater  thickness  of  the  same  liquid.  For 
the  explanation  of  this  property  of  the  copper  solution,  see  later 
(appendix). 

The  interposition  of  colored  cells  in  the  radiation  of  the  Carcel 

'"  I  have  many  times  used  the  light  obtained  in  this  way  for  works  of  astronomical 
photometry  (determination  of  the  candle-power  of  the  Sun,  comparison  of  the  light  of 
the  Sun  with  that  of  the  Stars,  etc.). 


3io 


TRANSACTIONS    I.  E.  S. — PART    II 


lamp  does  not  only  change  its  color ;  the  intensity  is  also  weak- 
ened. We  must  know  this  change  in  order  to  use  our  absorbing 
media  in  photometry.  The  weakening  is  partly  produced  by 
reflection  on  the  faces  of  the  cells,  and  partly  (the  greatest  part) 
by  absorption  in  the  liquids.  In  order  to  eliminate  the  effect  of 
reflection,  we  shall  compare  the  intensity  of  the  Carcel  through 
cells  filled  with  pure  water  with  the  intensity  through  the  same 
cells  containing  the  liquids ;  the  ratio  of  the  first  number  to  the 
second  will  be  called  the  opacity  of  the  liquids.  This  opacity  is 
a  perfectly  definite  function  of  the  two  quantities  x  and  y  which 
characterize  the  conditions  of  the  two  media.  It  is  possible  to 
investigate  it  once  for  all,  and  get  a  numerical  table  or  an  empi- 
rical formula  giving  the  values  of  the  opacities  for  every  value 


&5u  o.6u 

'WAVE- LENGTHS      ' 

Fig.  3- 

of  x  and  y.  The  experiment  necessary  to  get  the  numerical 
values  of  the  opacity  are,  unavoidably,  heterochromatic  photo- 
metric measurements ;  they  ought  to  be  made  with  the  most 
refined  methods  of  heterochromatic  photometry  and,  if  possible, 
by  several  observers,  to  have  the  most  probable  values  for  a 
normal  observer. 

I  made  some  such  measurements  in  1903;  they  were  not  pub- 
lished. I  will  give  here  my  results,  but  I  do  not  claim  for  them 
great  accuracy :  the  method  used  was  that  of  equal  brightness, 
with  the  use  of  the  Lummer-Brodhun  photometer  (the  flicker 
photometer  was  not  yet  in  general  use)  ;  perhaps  the  brightness 
of  the  photometric  screen  was  not  great  enough  to  eliminate  the 


Fabry:   problem  of  heterochromatic  photometry     311 

complications  arising  from  the  Purkinje  effect,  which  leads  to 
over-estimates  of  the  intensities  of  blue  lights ;  the  measurements 
were  made  only  by  one  observer.  I  have  some  reason  to  think 
that  the  values  given  for  opacity  of  the  blue  solution  were  rather 
too  low.  Be  that  as  it  may,  I  have  found  that  the  opacity  for 
thickness  x  and  y  of  the  two  luiquids  is  expressed  by  the  follow- 
ing empirical  formula :" 

log  opacity  =  0.016  x  -f  0.032  y  —  3  X  io-5  x-  +  ( 1 ) 

6  X  10-4  xy 
The  following  table  gives  some  values  computed   from  this 
equation : 


Opacity 

Opacity 

0 

0 

I. OO 

0 

1 

I.08 

10 

0 

1.44 

0 

2 

1. 16 

20 

0 

2.02 

0 

4 

1-34 

30 

0 

2.97 

0 

6 

1.56 

40 

0 

4.0 

10 

1 

1-57 

50 

0 

5-4 

20 

2 

2.49 

60 

0 

7-3 

40 

4 

6-55 

60 

6 

18.2 

The  values  of  the  opacity  being  known,  it  becomes  easy  to 
reduce  every  photometric  measurement  to  comparisons  of  the 
same  color.  Suppose  we  have,  for  instance,  to  measure  the 
intensity  of  an  electric  arc  in  terms  of  an  incandescent  carbon 
lamp  used  as  standard  of  intensity.  We  will  first  seek  what 
thickness  (or  concentration)  of  the  two  liquids  must  be  inter- 
posed in  front  of  the  Carcel  to  give  to  its  light  the  color  of  the 
electric  arc;  we  will  find  that  we  must  have  no  yellow  cell,  and, 
for  the  blue  one,  a  thickness  or  concentration  defined  by  x  =  33. 
On  the  other  hand,  the  incandescent  lamp  has  almost  exactly 
the  same  tint  as  the  Carcel ;  we  need  not  change  the  color  to 
compare  these  two  lights.  On  the  photometric  bench  we  put 
on  one  side  the  standard  lamp,  on  the  other  side  the  Carcel, 
interposing  the  cell  to  be  used  later,  filled  with  water ;  the  equal- 
ization of  the  two  intensities  will  be  made.  Instead  of  the 
standard  lamp  we  put  then  the  arc  to  be  measured,  and  we  fill 
the  cell  placed  in  front  of  the  Carcel  with  the  proper  blue  liquid  ; 
we  equalize  once  more  the  illuminations,  which  are  still  of  the 

11  For  the  derivation  of  this  formula  from  a  theoretical  standpoint  see  appendix. 
8 


312  TRANSACTIONS    I.  E.  S. — PART    II 

same  tint.     The  opacity  of  the  used  liquid  being  known,  the 
computation  of  the  intensity  of  the  arc  is  easy. 

This  procedure  is  scarcely  more  complicated  than  the  ordinary 
measurement  with  lights  of  the  same  color;  it  is  true  that  we 
must  equalize  twice  the  illuminations  on  the  photometric  screen, 
but  it  is  ever  so  when  the  double  weight  method  is  used  and 
the  use  of  this  method  is  generally  to  be  recommended.  If  we 
have  many  measurements  to  be  made  with  the  same  kind  of  light, 
the  same  liquid  will  be  used.  Perhaps,  in  this  case,  it  will  be 
found  more  convenient  to  use  some  colored  glass  properly  chosen, 
if  such  a  glass  can  be  found;  it  will  be  easy  to  measure  the  opac- 
ity of  this  glass  by  comparison  with  the  known  liquids,  without 
going  to  the  standardizing  bureau,  and  by  making  comparisons 
of  nothing  but  lights  of  the  same  color. 

It  is  worth  remarking  that  the  standard  color  lamp  (Carcel 
for  instance)  with  the  absorbing  liquids  gives  us  an  arbitrary 
scale  of  tints  whose  use  is  very  convenient  and  quite  inexpensive. 
True,  this  scale  is  quite  arbitrary,  but  from  the  data  so  obtained 
for  a  light,  it  is  always  possible  to  pass  to  the  tint  expressed  in 
absolute  value,  i.  e.,  the  values  of  fundamental  sensations  con- 
tained in  the  studied  light,  or  the  point  which  represents  it  on 
Maxwell's  diagram.  In  some  cases  where  the  other  methods  are 
difficult  to  use,  perhaps  this  procedure  could  give  some  useful 
information ;  it  would  be  perhaps  so  in  some  astronomical  ob- 
servations. 

As  was  said,  the  use  of  two  liquids  is  not  sufficient  to  produce 
every  possible  tint;  with  the  two  liquids  chosen,  we  can  only  get 
the  tints  represented  above  the  curve  AMB  (Fig.  i).  With  a 
third  liquid  we  could  get  every  possible  tint.  The  liquid  to  be 
chosen  ought  to  be  purple,  absorbing  the  middle  part  of  the' 
spectrum  more  than  the  two  ends;  this  medium,  taken  alone 
with  increasing  thickness  would  give  tints  represented  by  such 
a  curve  as  MC.  On  combining  this  liquid  with  the  blue  liquid 
A,  it  would  be  possible  to  have  every  tint  represented  by  points 
in  the  angle  AMC ;  the  same  liquid  in  combination  with  B  would 
allow  us  to  cover  the  angle  BMC.  In  short,  we  should  have 
three  liquids  giving  respectively,  when  employed  separately,  the 
curves  MA,  MB,  MC;  combinations  of  two  liquids  allow  the 


Fabry:    problem  of  heterochromatic  photometry     313 

production  of  every  possible  tint.     Perhaps  a  solution  of  potas- 
sium permanganate  would  be  good  as  a  third  liquid.12 

The  absorption  methods  in  optics  have  been  often  criticized, 
and  sometimes  rightly,  because  many  incorrect  results  have  been 
noticed  on  account  of  an  incorrect  interpretation  of  the  facts. 
Certainly,  the  use  of  absorption,  which  does  not  allow  the  com- 
plete separation  of  simple  radiations,  demands  a  critical  attitude 
on  the  part  of  the  inventor;  but,  when  the  methods  are  properly 
tried  out,  this  process  leads  to  very  simple  experimental  devices, 
which  is  of  first  importance  from  a  practical  standpoint. 

APPENDIX. 

I  will  collect  here  some  calculations  and  numerical  data  not 
necessary  for  the  practical  use  of  my  method,  but  useful  for  its 
comprehension. 

I  have  investigated  the  absorption  curve  (absorption  as  a  func- 
tion of  the  wave-length)  of  my  liquids.  The  measurements  were 
made  by  use  of  a  spectrophotometer,  without  seeking  for  a  great 
accuracy,  but  in  order  to  find  the  behavior  of  the  phenomenon. 

Let  us  consider  -the  liquid  A  ( 1  gram  crystallized  copper  sul- 
phate and  100  cubic  centimeters  ammonia  in  1  liter),  and  take 
it  with  a  thickness  of  1  millimeter  (.r  =  1).  For  a  simple  radia- 
tion of  wave-length  a,  this  cell  has  an  opacity  w  (ratio  of  the 
intensity  transmitted  through  water  and  intensit3r  through  the 
cell).  Instead  of  w  it  is  often  more  convenient  to  characterize 
the  absorption  by  the  decimal  logarithm  ot  w,  called  the  absorp- 
tion constant  a: 

a  =  log  opacity  =  absorption  constant. 
If  the  cell  has  a  thickness  x  the  value  of  the  absorption  constant 
for  the  same  wave-length  will  become  a.r. 

For  the  liquid  B  ( 1  gram  iodine  and  3  grams  potassium  iodide 
in  a  liter)  in  thickness  of  1  millimeter  (y  =  1)  the  similar  quan- 
tity will  be  denoted  /?. 

Fig.  4  gives  the  values  of  a  and  ft  plotted  against  the  wave- 
length. 

If  we  have  to  investigate  the  action  of- these  absorbing  media 
on  the  light  of  the  Carcel,  we  must  know  the  luminosity  curve 

12  The  use  of  the  third  liquid  would  probably  be  necessary  to  match  the  color  of  certain 
flaming  arc.    I  did  not  make  experiments  on  these  sources. 


3H 


TRANSACTIONS    I.  £.  S. — PART    II 


of  the  spectrum  of  its  light.  We  can  admit  that  the  radiation 
of  the  Carcel  lamp  is  identical  with  that  of  a  black  body  at  a 
temperature  a  little  under  2,000  degrees  of  the  absolute  scale; 
its  light  is  not  so  yellow  as  that  of  the  Carcel,  but  not  so  blue  as 
that  of  the  acetylene  flame.  We  can  therefore  trace  the  curve 
which  gives,  as  a  function  of  the  wave-length  A,  the  intensity 
I  measured  by  the  quantity  of  energy  carried  in  one  second 
(energy  curve  of  the  Carcel).  On  the  other  hand,  we  know  the 
sensibility  curve  of  the  eye  as  a  function  of  the  wave-length13 ; 
let  us  denote  by  8  the  sensibility  of  the  eye  for  the  radiation  A; 


0.6 

2 
O 

|0.5 
0 

§0.4 

I0-3 

0 

0 

"0.2 

0.1 

0.024 
0.022 

0.020§ 
F 

o.oia:> 

0.0 1&° 
111 

0.014  ji_ 

0.0 12  t 
o 

0.010% 

o 

0.0083 

0.006  ^ 

0.004 

0.002 


7r 


Fig.  4. 


the  values  of  S  are  proportional  to  the  luminous  intensity  at  equal 
energy  value.  The  product  L  =  18  is  the  luminous  intensity  (in 
arbitrary  units)  for  the  radiation  A,  and  we  can  trace  the 
luminosity  curve  of  the  spectrum  of  the  Carcel. 

We  put  in  front  of  the  Carcel  the  thickness  x  of  the  blue  liquid; 
the  opacity  of  this  cell  for  the  radiation  A  is  io"^;  this  radiation, 
whose  luminous  intensity  was  L,  will  be  reduced  to  I,  X  io— ax. 
It  will  be  easy  to  trace  the  luminosity  curve  of  the  light  so  modi- 
fied. 

In  order  to  have  the  opacity  of  the  cell  for  the  complete  light 
of  the  Carcel,  we  must  compare  the  total  intensities  without  and 
with  absorption.    In  the  first  case,  the  total  intensity  is : 

13  See,  for  instance :  Ives,  Transactions  of  Illuminating  Engineering  Society, 
April,  1911,  page  261,  and  October,  1912,  page  379;  or  Astrophysical  Journal,  November,  1912 


Fabry:    problem  of  heterochromatic  photometry     315 

With  absorbtion,  it  is: 

fLio-^A.. 

The  opacity  ft,  ratio  of  the  first  number  to  the  second,  is: 


a 


jW 


rdk 


A  similar  calulation  will  be  applied  to  the  liquid  B. 

If  we  use  the  two  liquids  A  and  B,  we  will  find  for  the  opacity: 

0  =  7 (2) 

jLicH^  +  fi-v)dK 

As  the  values  of  L,  a  and  /?  are  known  for  every  value  of  A,  we 
can  compute  the  values  of  the  opacity  for  every  value  of  x  and  y. 
The  integrals  are  computed  graphically,  and  that  is  not  difficult 
if  one  uses  a  planimeter. 

I  have  made  this  numerical  computation  for  some  values  of 
x  and  y.  The  concordance  with  experimental  values  given  above 
is  not  perfect,  but  tolerably  good,  in  consideration  of  the  lack  of 
accuracy  of  the  numerical  data  used  (luminosity  curve  of  the 
Carcel,  absorption  curve  of  the  liquids)  and  of  the  uncertainty 
of  my  heterochromatic  measurements. 

The  equation  2  which  gives  the  opacity  as  a  function  of  x  and  y 
is  not  expressible  by  the  elementary  functions.  We  can  express 
it  by  a  development  in  series.  The  mathematical  study  of  the 
question  shows  that  the  best  way  is  to  develop  log  O  in  increas- 
ing powers  of  x  and  y.  Dropping  powers  of  the  variables  higher 
as  the  second,  we  find : 

log  ft  =  Ax  +  By  +  C.v2  +  D/  +  Uxy. 

The  values  of  the  co-efficients  are: 


316 


TRANSACTIONS    I.  E.  S. — PART    II 


A  = 


J- 


hdX 


C  =  1,151 


D=  1,151 


3.302 


B     f 

\I<d\  ^d\ 

~(\phd\\        j/3'WA. 


(Jttft)'  jwx 


It  can  be  proved  that  A,  B  and  E  are  positive,  while  C  and  D 
are  negative.  The  formula  is  of  the  same  type  as  the  empirical 
equation  1 ;  the  computation  of  the  coefficients  by  the  above 
formula  gives  values  approximately  in  accordance  with  those 
empirically  obtained. 

Lastly,  we  can  compute  the  color  produced  by  a  given  thick- 
ness of  the  two  liquids  interposed  on  the  Carcel,  and  represent 
every  tint  by  a  point  on  Maxwell's  diagram.  In  this  connection, 
I  have  above  insisted  on  the  precision  with  which  the  blue  solu- 
tion alone,  interposed  in  front  of  the  Carcel,  can  change  its  color 
in  such  a  way  as  to  give  it  the  color  of  the  radiation  of  any 
black  body  at  higher  temperature.  To  explain  this  property,  we 
will  examine  the  following  problem : 

The  Carcel  is  supposed  to  have  the  same  radiation  as  a  black 
body  at  a  temperature  T  (a  little  under  2,000  degrees).  We  wish, 
by  a  properly  chosen  absorbant,  to  modify  its  radiation  in  such 
a  way  as  to  give  it  the  same  composition14  as  that  of  a  black  body 
at  a  temperature  T'  higher  than  T.  What  absorption  curve  must 
this  absorbent  have,  i.  e.,  what  is  the  relation  between  the  absorp- 
tion constant  of  the  cell  used  and  the  wave-length? 

14  That  is  a  sufficient  but  not  necessary  condition  to  have  the  desired  color.  If  it  is  ful- 
filled, the  tint  will  be  the  same  for  every  eye,  even  not  normal. 


Fabry:    problem  of  heterochromatic  photometry     317 

The  problem  is  immediately  solved  if  we  know  the  energy 
curve  of  the  two  radiations,  and  these  curves  are  denned  by 
Wien's  formula.15  The  opacity  of  the  cell  ought  to  be  the  ratio 
of  the  two  intensities,  multiplied  by  an  arbitrary  constant.  We 
find  thus  that  the  absorption  constant  a  (logarithm  of  opacity) 
must  be  expressed  as  a  function  of  the  wave-length  by  the  equa- 
tion : 

«  =  6,3.o(i-4;)!+C'<=-x+B. 
A  and  B  being  two  constants. 

If  we  trace  the  curve  giving  the  values  of  a  as  a  function  of 


-  we  must  have  a  straight  line.     If  such  a  substance  is  found. 

A 

it  will  allow  us  to  transform  the  radiation  of  a  black  body  at 
temperature  T  into  a  radiation  identical  with  that  of  a  black 
body  at  higher  temperature  T1 ;  the  thickness  of  absorbant  to  be 

used  is  proportional  to  -7=-  —  =7  . 

Returning  to  our  blue  liquid,  if  we  trace  the  curve  giving  its 

absorption  constant  as  a  function  of  —  ,  we  have  the  curve  in 

A 

Fig.  5.  That  is  not  mathematically  a  straight  line,  but  the  curve 
has,  in  the  brightest  part  of  the  spectrum,  a  very  long  inflection ; 
the  part  MN,  the  most  important  for  color,  is  almost  linear. 
It  is,  therefore,  not  surprising  that  the  color  (not  requiring  abso- 
lute identity  of  energy  curve)  of  the  black  body  at  any  high 
temperature  could  be  got  by  the  use  of  our  liquid.  In  fact,  the 
thickness  necessary  to  make  the  transformation  is  approximately 
given  by  the  equation : 


16,500  \y~  "TV  • 


The  determination  of  the  value  of  x  necessary  to  match  the 
color  of  the  Carcel  with  that  of  a  source  of  light  gives  us  an 
indication  of  the  temperature  of  this  source,  if  it  is  similar  to  a 
black  body. 

15  Planck's  equation  gives  practically  the  same  numerical  values  in  the  field  of  wave- 
length and  temperature  with  which  we  are  here  concerned. 


3i8 


TRANSACTIONS    I.  E.  S. PART    II 


The  black  body's  radiation  having  a  peculiar  importance  in  the 
science   of    illumination,    substances   having   an   absorption   law 

expressed  by  the  equation  a  = — |-  B  are  peculiarly  inter- 

esting  for  the  problem  of  heterochromatic  photometry.  Perhaps 
it  would  be  possible  to  find  some  other  liquid  fitting  this  law 
better  than  does  my  blue  solution,  but  that  is  a  matter  of  chance. 
Other  means  than  absorption  can  be  found  to  modify  the  intensi- 
ties of  different  radiations   (interference,  chromatic  or  rotatory 

0.024 

0.022 

0.020 

0.018 

0.016 

t  0.0 14 
o 

< 

o0.012 

o 

30.010 

&  0.008 

0.006 

0.004 

0.002 

0 

1.5  2  2.5^ 

Fig.  5. 

polarization),  and  they  have,  as  against  absorption,  this  ad- 
vantage that  the  law  of  alteration  can  be  computed  a  priori;  but 
it  does  not  seem  possible  to  get  by  one  of  these  methods  the 
law  of  opacity  expressed  by  the  above  equation.  The  only  com- 
plete solution  would  be  to  separate  the  different  simple  radiations 
in  a  spectrum,  then  weaken  these  differently  by  proper  screens, 
and  finally  recombine  them.  That  is,  of  course,  a  very  compli- 
cated procedure,  and  the  use  of  absorption,  in  spite  of  its  theo- 
retically imperfect  character,  seems  to  be  the  only  way  to  solve 
the  problem  practically. 


\M 

Ul 

0 

£ 

_z 

V 

1* 

1 

III 

z: 
■<. 

0 

Zj 

_J 

>- 

Ul 

Id 
ft: 
O 

—J 
CO 

0 

PROBLEM    OF   HETEROCHROMATIC    PHOTOMETRY  319 

DISCUSSION. 

Dr.  Herbert  E.  Ives  (Communicated)  :  The  Illuminating 
Engineering  Society  has  reason  to  feel  honored  by  Prof.  Fabry's 
selection  of  the  Transactions  for  the  place  of  publication  of 
this  admirable  paper.  His  contribution  is  marked  by  a  thorough 
grasp  of  the  subject,  by  much  insight  and  ingenuity  in  choice  of 
the  practical  means  for  carrying  out  the  central  idea  and  as  a 
whole  constitutes  a  very  positive  addition  to  the  progress  toward 
practical  colored  light  photometry. 

The  use  of  calibrated  absorbing  screens  has,  indeed,  as  Prof. 
Fabry  states,  been  suggested  before,  but  such  suggestions  have 
not  been  in  the  practical  shape  now  proposed  by  him.  They  have 
tacitly  involved  the  calibration  of  every  screen  sent  out  from  the 
standardizing  laboratory,  not  the  calibration  of  a  formula  easily 
reproducible  by  anyone.  This  constitutes  a  great  advance.  One 
cannot  but  admire,  too,  the  ingenious  idea  of  making  a  universal 
absorbing  screen  by  applying  the  three-color  principle  in  the 
choice  of  absorptions.  In  general  I  should  personally  have  been 
suspicious  of  this  method  through  fear  that  different  observers 
of  different  color  vision  would  not  find  the  standard  screens  to 
perform  their  color  difference  eliminating  function  completely. 
This  objection  is,  however,  very  fully  overcome  by  the  choice  for 
the  blue-green  element  of  a  medium  which  alone  is  sufficient  to 
perform  by  far  the  most  usual  transformation,  namely,  from 
one  black-body  color  to  another,  by  producing  very  nearly  an 
exact  spectral  match,  the  same  in  appearance  for  all  eyes. 

Up  to  the  publication  of  this  paper  it  has  happened  that  most 
attention  has  been  given,  by  workers  in  this  field,  to  study  aimed 
at  establishing  methods  for  calibrating  different  colored  standards. 
As  is  clearly  stated  by  Prof.  Fabry,  the  actual  practical  use  of 
his  absorbing  screens  is  dependent  upon  the  development  and 
adoption  of  such  methods  of  calibration.  In  view  of  the  recent 
progress  in  clearing  up  the  characteristics  of  different  photo- 
metric methods,  Prof.  Fabry's  screens  come  before  us  at  a  most 
opportune  time. 

It  may  be  of  interest  to  outline  here  the  work  along  this  line 
which  is  now  being  done  in  the  physical  laboratory  of  the  United 
Gas  Improvement  Company,  especially  as  the  new  absorbing 
9 


320  TRANSACTIONS    I.  E.  S. — PART    II 

screens  will  take  a  prominent  part  therein.  The  work  in  general 
consists  in  the  calibration,  by  several  different  methods,  of  stand- 
ards and  absorbing  screens  for  use  in  practical  photometry.  All 
of  these  methods  are  developments  of,  and  in  conformity  with, 
the  results  of  the  writer's  published  work  on  heterochromatic 
photometry.  In  short,  the  flicker  photometer,  under  the  condi- 
tions specified  in  the  writer's  paper  before  the  Society  at  the 
1912  convention,  will  be  the  standard  instrument,  and  the  visual 
luminosity  curve  determined  by  its  use  will  be  taken  to  represent 
the  normal  eye.  The  first  line  of  attack  is  through  the  use  of  a 
special  flicker  photometer,  now  under  construction,  embodying 
all  the  features  called  for  by  the  previously  mentioned  work. 
Auxiliary  standards  and  absorbing  screens  will  be  calibrated 
through  observations  by  a  large  number  of  observers.  The 
problem  of  selecting  absorbing  media  to  be  calibrated  is  obviously 
enormously  reduced  by  the  timely  appearance  of  Prof.  Fabry's 
work. 

A  second  line  of  attack  is  by  application  of  the  spectrum 
absorption  curve  of  the  various  media  to  the  normal  visual 
luminosity  curve  in  the  same  manner  that  Prof.  Fabry  has  indi- 
cated. Great  accuracy  is  aimed  at.  The  spectral  absorptions  will 
be  determined  by  the  aid  of  the  photo-electric  cell,  which  promises 
to  greatly  excel  the  eye  for  sensitiveness  in  spectrophotometry. 
In  all  cases  of  calibrating  screens  they  will  be  used  with  the 
standard  4-watt  lamp  as  maintained  at  the  Bureau  of  Standards, 
perhaps  also  with  the  standard  Pentane  lamp  and  the  Hefner. 

A  third  method  of  attack  is  through  the  photo-electric  cell  of 
one  of  the  alkali  metals,  properly  screened  so  as  to  have  the  sen- 
sibility curve  of  the  eye.  Experiments  toward  this  end  are  being 
actively  carried  on. 

It  is  only  in  one  line  of  work  that  calibrated  absorbing  screens 
might  be  found  inconvenient,  namely,  in  laboratory  tests  of  incan- 
descent mantles  of  varying  compositions  or  electric  incandescent 
lamps  of  varying  efficiency.  To  facilitate  this  kind  of  work  it 
is  my  present  idea  to  use  the  special  flicker  photometer  with 
absorbing  screens  which  will  correct  each  observer's  eye  to  nor- 
mal. Study  of  a  set  of  luminosity  curves  of  18  observers  has 
shown  that  practically  all  these  observers  can  be  corrected  to 


PROBLEM  OF  HETEROCHROMATIC  PHOTOMETRY     32 1 

normal  by  a  yellowish  or  bluish  screen  placed  over  the  eye. 
Given  two  colors  (such  as  solutions)  which  should  measure 
equally  bright  to  a  normal  observer,  and  two  solutions  showing 
absorption  curves  of  long  gradient,  it  should  be  possible  for 
every  observer  in  a  laboratory  to  be  equipped  with  a  screen  to 
carry  on  normal  heterochromatic  measurements. 

All  of  this  work  will  be  reported  in  due  course.  It  is  men- 
tioned here  in  order  to  indicate  the  prospects  that  before  long 
the  measurement  of  different  colored  lights  will  be  in  practical 
shape,  both  through  the  accurate  calibration  of  Prof.  Fabry's 
ingenious  screens  and  bv  other  means. 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 

Published  monthly,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIII 


OCTOBER.  1913 


No.  7 


Pittsburgh  Convention. 

The  seventh  annual  convention  of  the 
Illuminating  Engineering  Society  was 
held  at  the  Hotel  Schenley,  Pittsburgh. 
September  22  to  25,  1913.  A  program 
of  excellent  papers,  which  occasioned 
many  spirited  and  interesting  discus- 
sions, a  comparatively  large  attendance 
and  a  generous  complement  of  enter- 
tainment combined  to  make  this  meeting 
a  pronounced  success.  In  the  history  of 
the  society  it  marks  another  gain  in  the 
rapid  progress  made  by  the  society 
within  the  seven  years  of  its  existence 
in  the  promotion  of  the  science  and  art 
of  illumination. 

The  registration  totaled  459  members 
and  guests.  It  is  probable,  however,  that 
the  attendance  reached  approximately 
500.  Of  the  latter  number  176  were 
members.  The  average  attendance  at 
each  session  was  approximately  100. 

A  prize,  an  ornamental  desk  lamp,  was 
won  by  the  New  York  Section  for  hav- 
ing the  largest  representation  at  the 
meeting. 

Several  of  the  papers  and  reports 
appear  in  this  issue  of  the  Transac- 
tions; the  rest  will  be  published  in  the 
November  and   December   issues. 

A  brief  outline  of  the  convention  pro- 
ceedings is  given  in  the  following  para- 
graphs. 

Monday,  September  22. 

The  convention  was  called  to  order  at 


10  a.  m.  by  the  chairman  of  the  Conven- 
tion Committee,  Mr.  C.  A.  Littlefield. 
An  address  of  welcome  was  made  by 
Mr.  William  H.  Stevenson,  president  of 
the  Pittsburgh  Chamber  of  Commerce. 
An  appropriate  response  on  behalf  of 
the  society  was  made  by  Mr.  Norman 
Macbeth,  vice-president. 

A  historic  gavel  and  stand  was  pre- 
sented to  the  society  by  Prof.  George 
Hoadley  on  behalf  of  the  Philadelphia 
Section.  Inserted  in  the  gavel,  which  is 
of  rosewood,  are  several  contributions 
or  souvenirs,  each  one  of  which  desig- 
nates a  stage  or  significant  achievement 
in  the  lighting  industry.  The  stand  con- 
sists of  a  piece  of  a  gas  holder  built  in 
Baltimore  by  the  first  gas  company  in 
America,  established  in  1816.  set  in  a 
case  of  rosewood.  A  detailed  descrip- 
tion of  the  gavel  and  stand  will  appear 
in  another  issue  of  the  Transactions. 

The  address  by  President  Preston  S. 
Millar  surveyed  the  status  of  the  present 
day  lighting  conditions.  Based  on  an 
exhaustive  canvass  of  numerous  sources 
of  information,  it  constitutes  a  compre- 
hensive appraisal  with  which  future  esti- 
mates of  progress  in  the  science  and  art 
of  illumination  may  be  compared.  The 
address  will  be  published  in  a  later  issue 
of  the  Transactions. 

A  report  of  the  Committee  on  Organi- 
zation of  the  International  Committee 
on  Illumination  was  read  by  Mr.  George 
S.   Barrows.     The   report   reviewed   the 


TRANSACTIONS    I.    E.    S.—  PART    I 


movement  started  about  two  years  ago 
for  the  formation  of  an  international 
body  to  consider  questions  pertaining  to 
illumination.  A  meeting  was  held  July  8 
for  the  organization  of  a  United  States 
National  Committee  and  was  attended 
by  delegates  from  the  American  Gas 
Institute,  American  Institute  of  Elec- 
trical Engineers,  American  Physical 
Society  and  the  Illuminating  Engineer- 
ing Society.  Dr.  E.  P.  Hyde,  a  past- 
president  of  the  I.  E.  S.,  was  elected 
president  and  Dr.  C.  H.  Sharp  secretary 
of  the  United  States  National  Commit- 
tee. A  meeting  to  reorganize  the  Inter- 
national Photometric  Commission  was 
held  in  Berlin,  August  27  to  30,  1913. 
Ten  nations  were  represented  by  forty- 
four  delegates.  The  name  of  the  Com- 
mission was  changed  to  the  International 
Illumination  Commission  and  statutes 
for  its  conduct  were  adopted.  These 
provide  for  the  formation  of  national 
committees  by  the  national  technical 
societies  interested  in  lighting.  The 
business  of  the  Commission  is  to  be 
transacted  by  these  committees,  acting 
through  the  honorary  secretary  of  the 
Commission,  through  delegates  sent  to 
the  general  meetings  which  are  to  be 
held  every  three  years  and  through 
standing  committees.  The  next  meeting 
of  the  Commission  is  to  be  held  in  Paris 
in  1916.  A  more  detailed  and  formal 
report  on  the  proceedings  and  status  of 
the  new  Commission  is  to  be  received 
shortly  from  the  United  States  National 
Committee  and  will  be  published  in  a 
later  issue  of  the  Transactions. 

The  report  of  the  Committee  on 
Progress  was  next  presented  by  Mr. 
F.  N.  Morton,  chairman  of  the  com- 
mittee.   The  report  appears  in  this  issue. 

At  the  afternoon  session  three  papers 
were  presented :  "Recent  Improvements 
in  Incandescent  Lamp  Manufacture,"  by 
E.    J.    Edwards    and    Ward    Harrison; 


"The  Cooling  Effect  of  Leading  in 
Wires  upon  the  Filaments  of  Tungsten 
Incandescent  Lamps  of  the  Street  Series 
Type,"  by  T.  H.  Amrine;  "Modern 
Practise  in  Street  Railway  Illumination," 
by  S.  G.  Hibben. 

In  the  evening  there  was  a  reception 
and  dance  in  the  ball-room  of  the  Hotel 
Schenley. 

Tuesday,  September  23. 

Three  papers  were  read  during  the 
morning  session :  "The  Psychic  Values 
of  Light,  Shade,  Form  and  Color,"  by 
F.  Park  Lewis,  M.  D.;  "The  Efficiency 
of  the  Eye  Under  Different  Systems  of 
Illumination,  and  the  Effect  of  Varia- 
tions in  Distribution  and  Intensity  of 
Light,"  by  C.  E.  Ferree  of  Bryn  Mawr 
College;  and  "Some  Theoretical  Con- 
siderations of  Light  Production,"  by 
W.  A.  Darrah.  In  connection  with  the 
discussion  of  the  latter  paper,  Mr.  John 
W.  Howell  of  the  General  Electric  Com- 
pany gave  a  talk  on  current  develop- 
ments in  the  manufacture  of  incandescent 
lamps.  Several  large  high  candle-power 
nitrogen-filled,  tungsten  filament  lamps 
were  exhibited.  Mr.  C.  A.  B.  Halvor- 
son,  Jr.,  of  the  West  Lynn  Works  of  the 
General  Electric  Company  discussed  the 
status  of  the  arc  lamp,  referring  par- 
ticularly to  the  magnetite  lamp. 

In  the  afternoon  an  inspection  trip 
was  conducted  to  the  works  of  the 
Westinghouse  Electric  &  Manufacturing 
Company  in  East  Pittsburgh. 

The  Section  Development  Committee 
held  a  conference  of  the  representatives 
of  section  boards  for  the  discussion  of 
section  activities  for  the  season  begin- 
ning October  1. 

There  was  also  a  special  meeting  of 
the  Council  in  the  afternoon.  A  report 
of  this  meeting  appears  elsewhere  in 
this  issue. 


TRANSACTIONS    I.    E.    S.—  PART    I 


An  evening  session,  to  which  the 
public  was  invited,  was  held  in  the 
Allegheny  County  Soldiers'  Memorial 
Hall.  The  lighting  of  this  building, 
which  has  been  described  in  a  paper  in 
the  Transactions,  was  the  subject  of  a 
talk  by  Mr.  Henry  Hornbostel,  the 
architect  of  the  building. 

Mr.  Georges  Claude  gave  a  lecture  on 
the  neon  tube  lamp  01  which  he  is  the 
inventor,  and  showed  several  of  the 
lamps  in  operation.  The  lecture  is  in- 
cluded in  this  issue  of  the  Transac- 
tions. 

Two  papers,  "The  Evolution  of  the 
Lamp"  by  Roscoe  E.  Scott  and  "The 
Quartz  Mercury-vapor  Lamp''  by  W.  A. 
D.  Evans  were  also  presented. 

Wednesday,  September  24. 

This  day  was  designated  as  "Commer- 
cial Day"  and  was  devoted  to  a  dis- 
cussion of  various  lighting  installations 
involving  the  practical  applications  of 
scientific  lighting  principles.  The  papers 
discussed  were :  "Church  Lighting"  by 
R.  B.  Ely;  "Experiments  in  the  Illumi- 
nation of  a  Sunday  School  Room  with 
Gas"  by  E.  F.  Kingsbury;  "Distinctive 
Store  Lighting"  by  C.  L.  Law  and  A.  L. 
Powell ;  "Factory  Lighting"  by  M.  H. 
Flexner  and  A.  O.  Dicker;  "Store 
Lighting"  by  J.  E.  Philbrick;  "Hospital 
Lighting"  by  W.  S.  Kilmer. 

Wednesday  afternoon  the  members  of 
the  society  were  guests  at  a  baseball 
game  between  the  Pittsburgh  and 
Chicago  teams  of  the  National  League. 

The  annual  banquet  was  held  in  the 
evening.  Prof.  H.  S.  Hower  of  the 
Reception  Committee  presided  and  intro- 
duced the  toastmaster  General  George 
H.  Harries.  Addresses  were  made  by 
President  P.  S.  Millar,  President-elect 
C.  O.  Bond,  Charles  M.  Bregg  and  John 
Brashear,      "Pittsburgh's      Grand      Old 


Man."      Music    was    furnished    by    th> 
Westinghouse  Band. 

Thursday,  September  25. 

At  the  morning  session  the  followinj 
papers  were  read :  "The  Lighting  o 
Show  Windows"  by  H.  B.  Wheeler 
"The  Pentane  Lamp  as  a  Workins 
Standard"  by  E.  C.  Crittenden  and  A.  H 
Taylor;  "The  Illuminating  Engineering 
Laboratory  of  the  General  Electrii 
Company"  by  S.  L.  E.  Rose;  "Th< 
Photo-Electric  Cell  in  Photometry"  fy 
Prof.  F.  K.  Richtmyer;  "Some  Studie: 
in  the  Accuracy  of  Photometry"  by  E.  J 
Edwards  and  Ward  Harrison. 

At  noon  a  number  of  the  member; 
attended  a  luncheon  of  the  Jovians  ii 
the  Hotel  Schenley.  Statesman  F.  M 
Knapp  of  the  latter  organization  pre 
sided.  Short  talks  were  given  by  Pres 
ton  S.  Millar,  L.  B.  Marks,  Normar 
Macbeth,  C.  A.  Littlefield  and  Georg< 
Webster.  Music  was  furnished  by  th< 
orchestra  of  the  Western  Union  Tele- 
graph Company. 

The  afternoon  session  was  opened  bj 
a  paper  entitled  "Characteristics  of  En 
closing  Glassware"  by  V.  R.  Lansingh 
The  session  was  concluded  by  a  genera 
meeting  of  the  society.  The  annua 
report  of  the  Council  to  the  member 
ship  was  presented  by  Joseph  D.  Israel 
general  secretary.  The  report  will  b< 
published  in  a  later  issue  of  the  Trans 
actions. 

The  committee  appointed  at  the  firs 
session  of  the  convention  to  report  01 
the  presidential  address  then  presentee 
the  following  report : 

Your  Committee  desires  to  commend  the  Re 
port  of  your  President  for  the  careful  perusal  o 
each  individual  member  of  the  Society.  As  1 
record  of  the  present  status  of  the  Art  of  Ilium 
ination,  it  should  assume  an  important  place  ii 
the  archives  of  the  Society. 

The  classification  of  the  elements  of  lightin; 
practise  into  the  so-called  categories,  is  mos 
useful  not  only  in  expressing  present  day  con 


TRANSACTIONS    I.    E.    S. — PART    I 


ditions,  but  also  as  a  means  of  evaluating  future 
progress.  The  careful  and  extensive  survey  of 
lighting  conditions  through  broadcast  distribu- 
tion of  questions  will  undoubtedly  yield  much 
important  information. 

Your  Committee  trusts  that  the  President  will 
carry  this  investigation  through  to  a  conclusion 
and  make  available  such  instructive  data  as  may 
be  obtained. 

A  further  important  result  may  be  anticipated 
from  this  survey,  that  it  will  set  to  thinking  a 
large  number  of  persons  more  or  less  interested 
in  the  subject  of  Illumination,  thereby  stimulat- 
ing activity  on  the  subject. 

To  the  Council  it  is  desired  to  recommend  for 
consideration  the  President's  suggestion  that  an 
effort  be  made  to  secure  commercial  application 
of  the  valuable  information  which  is  being  con- 
tinuously  presented  to  the  Society. 

Your  Committee  believes  it  will  be  found  de- 
sirable to  appoint  a  committee  to  investigate  the 
suggestion  of  the  President,  and  recommend  not 
only  with  regard  to  carrying  it  out  but  also  with 
regard  to  the  ways  and  means. 

Respectfully  submitted  : 
G.  S.  Barrows, 
Norman  Macbeth, 
G.  H.  Stickney,  (Chairman). 

A  Committee  on  Resolutions  pre- 
sented the  following  resolutions,  which 
were  unanimously  adopted : 

Whereas,  The  Seventh  Annual  Convention  of 
the  Illuminating  Engineering  Society  is  about 
to  terminate  its  sessions,  and 

Whereas,  in  general  interest  it  has  fully 
maintained  the  high  standard  set  by  preceding 
meetings  of  the  Society,  and 

Whereas,  this  result  has  only  been  accom- 
plished through  the  indefatigable  effort  and  co- 
operation of  those  charged  with  the  responsi- 
bility of  the  preparation  for  and  the  conduct  of 
the  work.  Be  it  Resolved  : 

That  the  thanks  of  the  delegates  be  extended 

To  the  Pittsburgh  Section  for  their  invitation  to 
visit  Pittsburgh  and  their  generous  hospitality  ; 

To  the  General  Convention  Committee  for  the 
comprehensive  plan  laid  out  by  and  so  ably  ac- 
complished through  their  unceasing  efforts  ; 

To  the  various  Executive,  Local  and  other 
Committees  who,  by  unselfish  devotion  to  their 
duties,  have  so  largely  added  to  the  comfort  and 
entertainment  of  those  in  attendance  ; 

To  the  President  of  the  Pittsburgh  Chamber  of 
Commerce  for  his  cordial  welcome  ; 

To  the  County  Commissioners  for  permission 
to  use  Allegheny  County  Memorial  Hall  for  a 
session  ; 

To  the  Management  of  the  Hotel  Schenley  for 


the  many  courtesies  extended  to  the  delegates 
and  guests  ;  and 

To  the  authors  of  the  papers  and  the  speakers, 
without  whose  valued  efforts  the  success  of  this 
meeting  would  have  been  impossible  ;  and 

Be  it  Further  Resolved,  that  copies  of  these 
resolutions  be  forwarded  to  the  Pittsburgh  Sec- 
tion, to  the  President  of  the  Pittsburgh  Chamber 
of  Commerce,  to  the  County  Commissioners,  to 
the  Chairmen  of  the  various  Committees,  to  the 
Management  of  the  Hotel,  and  that  they  be 
spread  upon  the  minutes  of  the  Society. 

President  Millar  then  introduced 
President-elect  Charles  O.  Bond  who 
made  a  brief  address  in  response  to  the 
hearty  reception  which  was  accorded 
him. 

Mr.  L.  B.  Marks  introduced  a  resolu- 
tion commending  the  work  of  President 
Millar  and  his  administration,  which  was 
adopted  unanimously.  The  convention 
was  then  adjourned. 

In  the  evening  there  was  a  theater 
party  at  the  Grand  Opera  House. 

During  the  convention  special  enter- 
tainment was  provided  for  the  ladies. 
Monday  afternoon  there  was  a  card 
party  at  the  Pittsburgh  Athletic  Club. 
Tuesday  a  luncheon  was  served  at  the 
factory  of  the  H.  J.  Heinz  Company; 
there  was  also  an  automobile  trip 
through  the  city  parks.  Thursday  after- 
noon visits  were  made  to  the  Carnegie 
Institute  and  the  Margaret  Morrison 
School. 

In  the  way  of  sport  for  the  men  there 
were  several  games  of  golf,  and  a  tennis 
game  for  a  silver  trophy  which  was  won 
by  J.  L.  Wiltse. 


Council  Notes. 

During  the  seventh  annual  convention 
in  Pittsburgh,  a  special  meeting  of  the 
Council  was  held  September  23,  1913,  in 
the  Hotel  Schenley.  Those  present 
were:  Preston  S.  Millar,  president; 
George  S.  Barrows,  C.  O.  Bond.  Joseph 
D.  Israel,  general  secretary;  V.  R.  Lan- 


TRANSACTIONS    I.    E.    S.  —  PART    I 


singh,   Norman   Macbeth,   L.  B.   Marks, 
treasurer;  and  \V.  J.  Serrill. 

The  Executive  Committee  reported 
that  it  had  transacted  the  following  busi- 
ness since  the  last  regular  meeting  of 
the  Council  in  June.  The  report  was 
adopted. 

Authorized  the  payment  of  vouchers  Nos. 
1359  to  1390  inclusive,  amounting  to  $3600.48; 
and  vouchers  Xos.  1391  to  1420  inclusive, 
amounting  to  $1187.56. 

An  appropriation  of  $70  covering  the  cost 
of  programs,  postage,  etc.,  issued  in  conjunc- 
tion with  the  joint  meeting  of  the  Illuminating 
Engineering  Society  and  the  School  Hygiene 
Congress  in  Buffalo  (August  25-3°.  ,OI3),  was 
granted. 

Elected  twenty  applicants  to  membership. 
The  names  of  these  new  members  are  given 
on  another  page  of  this  issue  01  the  Trans- 
actions. 

Approved    the    appointment   of   the    following 
local   representatives: 
Abbott,   A.    L. 

Manager,    Northwestern    Electrical    Equip- 
ment  Company,    St.   Paul,    Minn. 
Collier,   William   Rawson. 

Georgia      Railway      &      Light      Company, 
Atlanta,   Ga. 
Hoar,  F.   Emerson. 

Railroad   Commission  of  the   State  of  Cali- 
fornia,  833   Market  Street,   San   Francisco, 
Cal. 
Manahan,  R.  H. 

City   Electrician,   Los  Angeles,   Cal. 
Osborn,  Fred.  A. 

University  of  \V  ashington,  Seattle,  Wash. 
Williamson,  G.  E. 

Denver  Gas  &  Electric  Light  Company, 
Denver,  Colo. 
Approved  appointment  of  two  representa- 
tives (Dr.  E.  P.  Hyde  and  Mr.  L.  B.  Marks) 
from  the  Illuminating  Engineering  Society  to 
a  meeting  of  representatives  of  several  socie- 
ties for  the  purpose  of  organizing  a  LTnited 
States  National  Committee.  The  organization 
of  this  committee  is  in  line  with  a  movement 
started  about  two  years  ago  for  the  formation 
of  an  international  body  to  consider  questions 
pertaining  to  illumination. 

A  draft  of  the  annual  report  of  the 
general  secretary  was  read.  This  report 
with  a  few  changes  was  made  the  report 


of  the  Council  to  the  membership  of  the 
society. 

An  appropriation  of  $1,005  was 
granted  to  cover  expenditures  to  be 
made  by  the  Convention  Committee  in 
connection  with  the  Seventh  Annual 
Convention. 

An  appropriation  of  $35  was  granted 
to  cover  the  cost  of  issuing  notices 
announcing  a  joint  meeting  of  the 
Illuminating  Engineering  Society  and 
the  American  Gas  Institute  during  the 
convention  of  the  latter  organization  in 
Richmond,  Ya.,  in  October. 

The  Finance  Committee  was  directed 
to  engage  a  public  accountant  to  audit 
the  accounts  of  the  society  for  the 
period  January  1  to  September  30,  1913, 
at  a  cost  not  to  exceed  $50. 

The  following  resolution  was  adopted. 

Whereas,  at  this  special  meeting  of 
the  Council  of  the  Illuminating  Engi- 
neering Society,  held  in  the  Hotel 
Schenley,  Pittsburgh,  Pa.,  September 
23,  1913,  during  the  period  of  the  Sev- 
enth Annual  Convention  of  the  Society, 
it  is  deemed  fitting  and  desirable  that 
the  Council  recognize  the  distinguished 
services  of  the  several  committees  which 
have  done  so  much  to  help  the  society's 
work  in  support  of  the  1913  administra- 
tion; therefore  be  it 

Resolved,  that  the  secretary  be  in- 
structed to  forward  to  the  chairman  of 
each  standing  and  temporary  committee, 
as  well  as  to  the  chairmen  of  the  con- 
vention and  sub-committees  a  copy  of 
this  resolution,  as  an  appreciation  by  the 
Council  of  the  committee  work  which 
was  so  marked  a  feature  of  the  activi- 
ties of  the  society  during  the  year  1913. 
and  as  a  recognition  of  the  efficiency 
with  which  the  several  committees  per- 
formed their  various  duties;  be  it 
further 

Resolved,  that  the  Coun:il  place  on 
record  in  this  manner  a  renewal  of  its 


TRANSACTIONS    I.    E.    S. — PART    I 


expression  of  appreciation  of  the  fidelity 
and  efficiency  of  the  Assistant  Secretary, 
Mr.  Joseph  Langan,  and  his  office  assist- 
ants. 

It  was  voted  to  recommend  to  the  next 
Council  the  question  of  considering 
ways  and  means  of  establishing  the 
responsibility  of  the  society  for  all  state- 
ments made  in  papers  presented  by  mem- 
bers of  the  society  at  joint  meetings  with 
other  societies. 

Prof.  George  D.  Shepardson  of  the 
University  of  Minnesota,  Minneapolis, 
Minn.,  was  appointed  a  local  represen- 
tative of  the  society. 

Mr.  V.  R.  Lansingh,  chairman  of  the 
Committee  on  Sustaining  Membership, 
reported  on  the  status  of  the  sustaining 
membership  of  the  society. 

It  was  voted  to  refer  to  the  next 
Council  the  question  of  supplying  re- 
prints and  copies  of  papers  to  authors. 

Vouchers  Nos.  1421  to  1442  inclusive, 
amounting  to  $691.20,  were  authorized 
paid,  subject  to  subsequent  approval  by 
the  Finance  Committee. 

The  first  regular  Council  meeting  of 
the  administration  1913-1914  was  held  in 
the  general  offices  of  the  society,  29  West 
39th  Street,  New  York,  October  10,  1913. 
In  attendance  were:  C.  O.  Bond,  presi- 
dent; Ward  Harrison,  J.  D.  Israel,  gen- 
eral secretary;  V.  R.  Lansingh,  C.  A. 
Littlefield,  L.  B.  Marks,  treasurer; 
Preston  S.  Millar,  W.  J.  Serrill,  G.  H. 
Stickney. 

The  meeting  was  called  to  order  by 
President  Bond  at  10:40  a.  m. 

The  minutes  of  a  special  meeting  of 
the  Council,  which  was  held  September 
23  in  the  Hotel  Schenley,  Pittsburgh, 
Pa.,  during  the  Annual  Convention, 
were  adopted. 

Payment  of  vouchers  1443  to  1458  in- 
clusive, amounting  to  $992.10,  was 
authorized. 

Ten  applicants  were  elected  members. 


Their  names  appear  elsewhere  in  this 
issue. 

Eleven  resignations  from  membership 
were  accepted. 

Two  applicants  for  sustaining  mem- 
bership were  elected. 

It  was  voted  to  hold  the  regular  meet- 
ings of  the  Council  during  the  present 
administration  on  the  Friday  following 
the  second  Thursday  of  each  month,  at 
10:30  a.  m. 

The  following  suggestions  from  the 
preceding  Council  were  referred  to  the 
Finance  Committee  for  recommenda- 
tions : 

(1)  Set  aside  at  the  beginning  of  each  year 
for  immediate  investment  in  such  securities  as 
may  be  recommended  by  the  Finance  Com- 
mittee, a  sum  of  money  equal  to  3  per  cent, 
of  the  preceding  year's  total  revenue.  To 
provide  for  future  contingencies  of  an  extreme 
character,  some  such  provision  seems  war- 
ranted. 

(2)  The  moneys  of  the  Society  might  be 
deposited  in  a  bank  or  trust  company  which 
will  allow  interest  on  monthly  or  periodical 
balances.  The  revenue  from  such  a  source 
might  easily  exceed  $100  annually. 

The  following  suggestion  from  the 
preceding  Council  was  discussed  and  re- 
ferred to  the  Committee  on  Progress  for 
recommendations : 

With  a  view  to  making  the  Transactions 
more  valuable  as  a  reference  publication,  a 
list  of  current  books  and  articles  pertaining 
to  illuminating  engineering  might  be  published 
each  month  in  the  Transactions.  Were  a  brief 
statement  of  the  contents  to  accompany  the 
titles  of  these  publications,  this  contribution 
would  be  of  so  much  greater  value. 

President  Bond  was  directed  to  ap- 
point a  special  committee  to  reconsider 
the  policy  of  the  society  in  regard  to 
supplying  authors  with  copies  of  their 
papers  which  are  published  in  the 
Transactions.  The  present  practise  is 
to  give  each  author  50  of  the  advance 
copies  of  his  paper.  When  a  paper  is 
not  published  in  advance  form,  the 
author    does    not    receive    any    copies. 


TRANSACTIONS    I.    E.    S. — PART    I 


President  Bond  appointed  the  following 
committee:  G.  H.  Stickney,  Chairman; 
Dr.  C.  H.  Sharp,  Dr.  Herbert  E.  Ives. 
It  was  understood  that  the  committee 
would  submit  recommendations  at  the 
November  meeting  of  the  Council. 

The  question  of  establishing  the  re- 
sponsibility of  the  society  for  statements 
made  in  papers  presented  by  members 
of  the  society  at  joint  meetings  with 
other  societies,  and  before  meetings  of 
the  Illuminating  Engineering  Society, 
was  referred  to  the  Papers  Committee 
for  recommendations,  to  be  submitted  at 
the  November  meeting  of  the  Council. 
It  was  suggested  that  a  note  of  some 
sort  might  be  placed  upon  each  paper, 
stating  that  the  society  does  not  neces- 
sarily identify  itself  with  the  views  or 
opinions  expressed  in  its  papers. 

An  invitation  from  the  American 
Museum  of  Safety  to  the  society,  to  join 
in  a  session  to  be  devoted  to  illumina- 
tion, during  an  exhibition  of  the  former 
organization  in  New  York,  was  accepted. 
It  was  understood  that  the  Papers  Com- 
mittee and  the  Committee  on  Reciprocal 
Relations  conjointly  would  designate 
lecturers  or  authors  to  appear  at  the 
session  of  the  American  Museum  of 
Safety,  on  the  evening  of  December  18 
in  the  Grand  Central  Palace,  New  York 
City. 

In  connection  with  the  latter  meet- 
ing, $25  was  appropriated  for  litera- 
ture to  be  issued  in  conjunction 
with  this  session  and  a  lighting  exhibit 
which  is  to  be  installed  in  the  Museum 
of  Safety  by  a  committee  of  the  society. 
It  was  understood  that  the  exhibit  would 
also  be  displayed  in  the  Grand  Central 
Palace  during  the  Museum  of  Safety 
Exhibition.  This  exhibit  is  to  become 
part  of  the  permanent  equipment  of  the 
Museum  of  Safety. 

Mr.  G.  H.  Stickney,  chairman  of  the 
committee  having  charge  of  the  installa- 


tion of  a  lighting  exhibit  in  the  Ameri- 
can Museum  of  Safety,  reported  briefly 
on  the  work  of  his  committee.  He 
stated  that  the  expense  of  construction 
and  installation  of  the  exhibit  would  be 
borne  by  several  lighting  and  manufac- 
turing companies. 

It  was  voted  to  donate  to  the  School 
Hygiene  Congress  a  set  of  electros  of 
the  illustrations  in  the  illumination 
primer  "Light:  Its  Use  and  Misuse"  and 
a  copy  of  the  primer,  with  the  under- 
standing that  it  would  be  incorporated 
verbatim  and  in  its  entirety  in  the  pub- 
lished proceedings  of  the  Congress 
which  was  held  in  Buffalo  in  August, 
1913.  The  electrotypes  are  to  be  for- 
warded to  Dr.  Thomas  A.  Storey  of  the 
College  of  the  City  of  New  York,  who 
is  secretary  of  the  Congress. 

Mr.  Ward  Harrison  of  the  National 
Electric  Lamp  Association  of  Cleveland, 
was  appointed  a  local  representative  of 
the  society  in  the  city  of  Cleveland,  Ohio. 
The  following  local  representatives 
were  reappointed  for  the  present 
administration : 
Abbott,  A.  L. 

Manager,     Northwestern     Electrical 
Equipment  Company,  St.  Paul,  Minn. 
Collier,  William  Rawson. 

Georgia  Railway  &  Light  Company, 
Atlanta,  Ga. 
Hoar,  F.  Emerson. 

Railroad   Commission   of   the   State 
of    California,    833    Market    Street, 
San  Francisco,  Cal. 
Manahan,  R.  H. 

City  Electrician,  Los  Angeles,  Cal. 
Osborx,  Fred.  A. 

University  of   Washington,   Seattle, 
Wash. 
Shepardson,  G.  D. 

Professor  of  Electrical  Engineering, 
University  of  Minnesota,  Minneapo- 
lis, Minn. 


8 


TRANSACTIONS   I.    E.    S. — PART    I 


Williamson,  G.  E. 

Denver  Gas  &  Electric  Light  Com- 
pany, Denver,  Colo. 

The    following    committee    chairmen 
were  appointed : 

Progress : 

F.  E.  Cady,  National  Electric  Lamp 
Association,  Cleveland,  Ohio. 

Glare : 

M.      Luckiesh,      National      Electric 
Lamp  Association,  Cleveland,  Ohio. 

Popular  Lectures: 

G.  H.  -Stickney,    General    Electric 
Company,  Harrison,  N.  J. 

Sustaining  Membership  : 

V.  R.  Lansingh,  6523  Euclid  Avenue, 
Cleveland,  Ohio. 

Advertising  : 

J.     Robert     Crouse,    The    National 

Electric     Lamp     Association,     Nela 

Park,  Cleveland,  Ohio. 
Lighting  Legislation  : 

L.    B.    Marks,    103    Park    Avenue. 

New  York  City. 
Section  Development: 

Joseph     D.     Israel,     1000     Chestnut 

Street,  Philadelphia,  Pa. 

Papers  : 

George  H.  Stickney,  General  Elec- 
tric Company,  Harrison,  N.  J. 
It  was  resolved  that  the  Council  at 
this  time  renew  its  expression  of  appre- 
ciation of  the  excellent  work  done  by 
the  Primer  Committee  in  the  prepara- 
tion of  the  illumination  primer  "Light : 
Its  Use  and  Misuse" — the  best  piece  of 
constructive  work  the  society  has  ever 
accomplished. 

The     following     Council     Executive 
Committee  was  appointed: 
Charles   O.   Bond    (ex-Ofiicio),   Chair- 
man, 
3101  Passyunk  Avenue,  Philadelphia, 
Pa. 


Joseph  D.  Israel  (ex-OMcio), 

1000   Chestnut   Street,    Philadelphia, 

Pa. 
C.  A.  Littleeield, 

55  Duane  Street,  New  York  City. 
L.  B.  Marks  (ex-OMcio), 

103  Park  Avenue,  New  York  City. 
Preston  S.  Millar, 

80th  Street  and  East  End  Avenue, 

New  York  City. 


Section  Notes. 

CHICAGO    SECTION. 

A  meeting  of  the  Board  of  Managers 
was  held  in  the  Grand  Pacific  Hotel, 
October  15.  Those  present  were :  Dr. 
M.  G.  Lloyd,  chairman ;  J.  R.  Cravath, 
vice-president;  J.  B.  Jackson,  secretary; 
J.  W.  Pfeifer,  C.  C.  Schiller,  and  H.  B. 
Wheeler. 

The  getting  of  suitable  papers  on  car 
lighting  for  the  November  meeting  was 
discussed.  Dr.  Lloyd  and  Mr.  Jackson 
were  delegated  to  complete  arrange- 
ments for  the  meeting. 

It  was  announced  that  Mr.  W.  A. 
Durgin,  assistant  chief  testing  engineer 
of  the  Commonwealth  Edison  Company, 
would  give  a  series  of  talks  on  the  basic 
principles  of  illumination  before  the  sec- 
tion meetings  during  the  present  season. 

A  proposed  joint  meeting  with  the 
Railway  Signal  Association  was  dis- 
cussed. The  making  of  definite  arrange- 
ments for  this  meeting  was  withheld 
pending  further  correspondence  with  the 
association. 

Committee  appointments  will  be  an- 
nounced later. 

A  meeting  of  the  Chicago  Section  was 
held  in  the  auditorium  of  the  Western 
Society  of  Engineers,  Monadnock  Block, 
October  15.  Vice-president  J.  R.  Cra- 
vath reviewed  the  proceedings  of  the 
Pittsburgh    convention.      Two    papers : 


TRANSACTIONS   I.    E.    S. — PART    I 


"Factory  Lighting"  by  Messrs.  M.  H. 
Flexner  and  A.  O.  Dicker  and  "The 
Lighting  of  Show  Windows"  by  H.  B. 
Wheeler  were  presented.  Demonstra- 
tions and  exhibits  of  factory  lighting 
fixtures,  show  window  reflectors  and  the 
new  nitrogen  lamps  were  provided.  A 
report  from  the  previous  board  on  the 
work  of  the  past  year  was  read  by  Sec- 
retary J.  B.  Jackson.  Dr.  M.  G.  Lloyd 
presented  a  tentative  outline  of  the 
meetings,  papers,  etc.,  for  the  present 
year.  Twenty-five  members  and  five 
guests  were  present. 

NEW    ENGLAND    SECTION. 

The  Board  of  Managers  of  the  New 
England  Section  expect  to  announce 
shortly  a  tentative  program  of  papers 
and  meetings  for  the  present  season. 

Mr.  C.  M.  Cole,  156  Pearl  Street, 
Boston,  Mass.,  has  been  appointed  sec- 
retary of  the  section. 

NEW   YORK    SECTION. 

The  Board  of  Managers  held  two 
meetings  in  the  general  offices  of  the 
society,  29  West  39th  Street,  New  York, 
September  2  and  29  respectively.  A 
brief  of  the  business  transacted  at  these 
meetings  is  given  below : 

It  was  decided  that,  whenever  possible 
during  the  present  year,  joint  meetings 
should  be  held  with  other  societies. 
Meetings  with  the  following  organiza- 
tions were  suggested  :  National  Electric 
Light  Association,  Sage  Foundation  for 
the  Prevention  of  Blindness,  American 
Society  of  Mechanical  Engineers,  New 
York  Electrical  Society,  National  Com- 
mercial Gas  Association  and  the  Munici- 
pal Art  Society.  The  secretary  was  in- 
structed to  prepare  a  budget  for  the 
coming  season.  An  effort  will  be  made 
to  have  issued  some  time  in  October  or 
early   in   November  a   printed   tentative 


program,  listing  papers,  meetings,  etc., 
for  the  entire  season.  Mr.  A.  S.  Ives 
was  appointed  a  manager  to  succeed 
C.  R.  Clifford,  resigned.  It  was  voted 
to  hold,  as  far  as  possible,  an  informal 
dinner,  immediately  preceding  each  sec- 
tion meeting,  at  Keene's  Chop  House, 
70  West  36th  Street,  New  York,  on 
the  evenings  of  the  section  meetings. 
Twelve  applications  for  membership 
were  approved  by  the  Board  of  Mana- 
gers and  transmitted  to  the  secretary  of 
the  society. 

The  following  committee  chairmen 
have  been  appointed :  Papers,  H.  V. 
Allen;  Membership,  H.  B.  McLean; 
Dinner,  S.  W.  Van  Rensselaer ;  Attend- 
ance, N.  D.  MacDonald ;  Exhibition, 
W.  H.  Spencer ;  Reception,  C.  L.  Law. 

A  meeting  of  the  New  York  Section 
was  held  in  the  United  Engineering 
Societies'  Building,  29  West  39th  Street, 
New  York,  October  9.  Mr.  Norman 
Macbeth  reviewed  the  proceedings  of 
the  seventh  annual  convention  of  the 
society  in  Pittsburgh,  in  September. 

A  paper  entitled  "Distinctive  Store 
Lighting,"  which  had  been  presented  at 
the  convention,  was  presented  by  Messrs. 
C.  L.  Law  and  A.  L.  Powell.  The  latter 
paper  was  accompanied  by  a  series  of 
lantern  slides  and  autochrome  color 
plates  demonstrating  the  advantages  of 
using  color  slides  in  photographic  work. 
Preceding  the  meeting  there  was  an  in- 
formal dinner  at  Keene's  Chop  House, 
70  Wrest  36th  Street. 


PHILADELPHIA    SECTION. 

The  Board  of  Managers  of  the  Phila- 
delphia Section  held  two  meetings,  one 
July  i&  and  another  September  5,  1913. 

The  following  committee  chairmen 
have  been  appointed :  Papers,  H.  A. 
Hornor;  Exhibition,  H.  Calvert;  Mem- 
bership,   Samuel    Snyder;    Publicity,    H. 


10 


TRANSACTIONS    I.    E.    S. — PART    I 


H.  Ganser;  Dinner,  F.  C.  Dickey; 
Attendance,  R.  B.  Ely. 

A  meeting  of  the  Philadelphia  Section 
was  held  in  the  Engineers'  Club,  1317 
Spruce  Street,  Philadelphia,  October  17. 
A  paper  entitled  "A  Simple  Unit  Method 
of  Measuring  Intrinsic  Actinicity  of 
Flames  and  Surfaces  for  the  Practise 
of  Photography"  by  Mr.  F.  M.  Steadman 
was  presented  by  Prof.  A.  W.  Good- 
speed,  of  the  University  of  Pennsylva- 
nia. Vice-President  W.  J.  Serrill  re- 
viewed the  proceedings  of  the  seventh 
annual  convention  of  the  society  in 
Pittsburgh,  September,  1913.  During 
the  meeting  there  was  an  exhibition  of 
photographic  lamps  and  lenses.  About 
60  members  were  present.  An  informal 
dinner  at  the  Engineers'  Club  preceded 
the  meeting. 

The  following  program  of  meetings 
and  papers  has  been  printed  and  issued 
to  the  members  of  the  section. 

Friday,  October  17. 
Joint  Meeting  with   Photographic 
Society  of  Philadelphia. 
"A  Simple  Unit  Method  for  Measuring 
the    Intrinsic   Actinicity   of    Flames 
and    Surfaces    for    the    Practise    of 
Photography."    By  Mr.  F.  M.  Stead- 
man. 

Presented  by  Prof.  A.  W.  Goodspeed. 
Address  by  Prof.  Geo.  A.  Hoadley. 
Report  of  Pittsburgh  Convention. 

By  Mr.  William  J.  Serrill. 
Exhibition  of  Photographic  Lenses  and 
Shutters. 

Thursday,  November  20. 

Joint  Meeting  with  Ophthalmologists  of 

Philadelphia   at  the   College   of 

Physicians  and  Surgeons. 

"Some  Effects  of  Artificial  Light  upon 

the  Eye." 

By  Dr.  Edward  A.  Shumway. 


"The    Effect    of    Mercury    Arc    Lamps 
upon  the  Eyes." 

By  Dr.  Geo.  S.  Crampton. 
"The  Lighting  of  a  Private  Library." 

By  Prof.  Arthur  J.  Rowland. 

Monday,  December  8. 
Joint  Meeting  with  Philadelphia  Section 

A.  I.  E.  E. 
"Brightness   Measurements   versus   Illu- 
mination Measurements." 

By  Dr.  Herbert  E.  Ives. 
"Railway  Car  Lighting." 

By  Mr.  Geo.  H.  Hulse. 
"The  Mercury  Quartz  Tube  Lamp." 

By  Mr.  Buckman. 

Friday,  January  6. 
"Deficiencies  of  the  Method  of  Flicker 
for    the    Photometry    of    Lights    of 
Different  Colors." 

By  Prof.  C.  E.  Ferree. 

Saturday,  February  7. 
Meeting  under  the  Auspices  of  Drexel 

Institute. 
"Light  and  How  to  Use  It." 

By  Mr.   C.  O.  Bond,   President 
of  I.  E.  S. 

Wednesday,   February   18. 
Joint  Meeting  with  Franklin  Institute. 
"Artificial  Daylight." 

By  Dr.  Herbert  E.  Ives. 

Friday,  March  20. 
"Lighting    and    Signalling    Systems    of 
Subways." 

By  Mr.  F.  D.  Bartlett. 
"The  Sun — The  Master  Lamp." 

By  Prof.  James  Barnes. 

Thursday,  April  9. 
Joint  Meeting  with  Franklin  Institute. 
"Recent    Developments    in    the    Art    of 
Illumination." 

By  Mr.  Preston  S.  Millar. 


TRANSACTIONS    I.    E.    S.  —  PART    I 


11 


Friday,  April  17. 
"The  Structure  of  the  Normal  Eye  and 
its  Ability  to  Protect  Itself  Against 
Ordinary  Light." 

By  Dr.  Wendell  Reber. 
"Glassware  for  Illumination  and  Other 
Purposes." 

By  Mr.  James  Gillinder. 

Friday,  May  15. 

Mass    Meeting    of    all    the    Engineering 

Societies  of   Philadelphia  and 

Vicinity. 

Special  Program  to  be  arranged  and  to 

include  an  address  on 
"The     Relation     of    Engineers    to     the 
Progress  of  Civilization." 

By  Dr.  Chas.  Proteus  Steinmetz. 


PITTSBURGH   SECTION. 

The  Board  of  Managers  of  the  Pitts- 
bugh  Section  has  announced  the  follow- 
ing tentative  program  of  meetings  and 
papers  for  the  present  season : 
October — "Technical    Elements    of    Vi- 
sion" by  Dr.  H.  H.  Turner. 
November — -"Technical      Discussion     of 
the  Elements  of  Lighting"  by  Prof. 
Hower  and  others. 
December — A    joint    meeting    with    the 
Pittsburgh  Section  of  the  American 
Institute  of  Electrical  Engineers.    A 
Central   Station   paper  will  be  pre- 
sented   by    H.    N.    Muller    of    the 
Duquesne  Light  Company. 
January — A  paper  to  be  selected  by  the 
members  from  Cleveland.    The  sub- 
ject will  be  announced  later. 
February — "Railroad  Yard  Lighting"  by 
A.  C.  Cotton  and  A  Kirschberg  of 
the     Pennsylvania     Railroad     Com- 
pany. 
March — A   "gas     lighting     subject;     the 

speaker  to  be  announced  later. 
April — "Developments  of  Flame  Carbon 
Arc  Lamps"  by  C.  E.  Stephens. 


May — "Store  Lighting" ;    speaker   to   be 

announced  later. 
June — Open. 

The  following  committee  chairmen 
have  been  appointed :  Papers,  C.  E. 
Stephens ;  Exhibition,  S.  G.  Hibben ; 
Membership,  C.  H.  Mohr;  Publicity, 
M.  C.  Turpin;  Dinner,  R.  H.  Skinner; 
Reception,  V.  R.  Lansingh;  Education, 
W.  A.  Darrah. 


New  Members. 

The  following  applicants  were  elected 

members    of    the    society    September 

1913: 

Airston,  Alexander. 

Head  of  Factory  Lighting  Depart- 
ment, Westinghouse  Electric  &  Mfg. 
Company,  East  Pittsburgh,  Pa. 

Blumenauer,  C.  H. 

President  and  Treasurer,  Jefferson 
Glass  Company,  Follansbee,  W.  Va. 

Bowen,  Dudley  A. 

Salesman,  Westinghouse  Electric  & 
Mfg.  Company,  165  Broadway,  New 
York,  N.  Y. 

Brand,  Walter  C. 

Illuminating  Engineer,  Macbeth- 
Evans  Glass  Company,  Wabash 
Building,  Pittsburgh,  Pa. 

Bullard,  John  E. 

Manager  New  Business  Department, 
Public  Service  Corporation  of  L.  I., 
Port  Washington,  N.  Y. 

Cadby,  J.  N. 

Chief  Inspector  of  Electric  Service, 
R.  R.  Commission  of  Wisconsin, 
Madison,  Wis. 

Damron,  C.  E. 

Assistant  in  Commercial  Engineer- 
ing Department,  General  Electric 
Company,  Harrison,  N.  J. 

Hake,  Harry  G. 

Assistant  Professor  of  Electrical 
Engineering,  Washington  Univer- 
sity, St.  Louis,  Mo. 


12 


TRANSACTIONS    I.    E.    S. — PART    I 


Harrison,  Haydn  T. 

Electrical  and  Illuminating  Engi- 
neer, ii  Victoria  Street,  London, 
S.  W.,  England. 

Langworthy,  E.  L. 

Eastern  Manager,  The  Adams  & 
Westlake  Company,  2218  Ontario 
Street,  Philadelphia,  Pa. 

McNary,  S.  J. 

Vice-President  and  Illuminating  En- 
gineer, The  E.  G.  Jones  Electric 
Company,  141  East  4th  Street,  Cin- 
cinnati, Ohio.  lyS 

Pascoe,  C.  C. 

Manager  of  Lamp  Sales  and  Illumi- 
nation Department,  General  Electric 
Company,  185  Paseo  Colon,  Buenos 
Aires,  Argentine  Republic. 

RilEy,  F.  M.  H. 

Lighting  Specialist,  Kansas  City 
Electric  Light  Company,  Grand 
Avenue,  Kansas  City,  Mo. 

Stannard,  Clare  N. 

Secretary,  Denver  Gas  &  Electric 
Light  Company,  Denver,  Colo. 

Swallow,  Joseph  G. 

Chief  of  Installation  and  Inspection, 
United  Electric  Light  &  Power 
Company,  11 70  Broadway,  New 
York,  N.  Y. 

Trotter,  A.  P. 

Electrical  Engineer  and  Advisor, 
Board  of  Trade,  Whitehall,  London, 
S.  W.,  England. 

Voyer,  Leonard  E. 

Assistant  in  Illuminating  Engineer- 
ing Laboratory,  General  Electric 
Company,  Harrison,  N.  J. 

Walters,  G.  L. 

Sales  Manager,  The  Adams  &  West- 
lake  Company,  319  West  Ontario 
Street,  Chicago,  111. 

Watkins,  Frederick  A. 

District  Sales  Manager,  Pittsburgh 
Reflector  &  Illuminating  Company, 
565  West  Washington  Street,  Chi- 
cago, 111. 


Williamson,  G.  E. 

Denver  Gas  &  Electric  Light  Com- 
pany, Denver,  Colo. 


The  following  applicants  were  elected 
members  October  10,  1913  : 

AtmorE,  A.  L. 

Engineering  Department,  Philadel- 
phia Electric  Company,  1000  Chest- 
nut Street,  Philadelphia,  Pa. 

Barnes,  James. 

Professor  of  Physics,  Bryn  Mawr 
College,  Bryn  Mawr,  Pa. 

Federer,  Theo.  P. 

Incandescent  Lamp  Salesman,  Gen- 
eral Electric  Company,  30  Church 
Street,  New  York,  N.  Y. 

Fortune,  F.  R. 

Lighting  Specialties  Company,  650 
Century  Building,  Pittsburgh,  Pa. 

Gray,  Henry  J.  W. 

Philadelphia  Electric  Company,  1000 
Chestnut  Street,  Philadelphia,  Pa. 

Loeouist,  Harry  S. 

Engineering  Department,  National 
Lamp  Works  of  General  Electric 
Company,  Nela  Park,  Cleveland,  O. 

McQuiston.  J.  C. 

Advertising  Manager,  Westinghouse 
Electric  &  Mfg.  Company,  East 
Pittsburgh,  Pa. 

Moon,  T.  Elmer. 

Illuminating  Engineer,  James  '  E. 
Hamilton  Engineering  Co.,  1625 
Real  Estate  Trust  Building,  Phila- 
delphia, Pa. 

Strait,  E.  N. 

Public  Utility  Expert,  Railroad 
Commission  of  Wisconsin,  Madison, 
Wis. 

Turpin,  M.  C. 

Westinghouse  Electric  &  Mfg.  Com- 
pany, East  Pittsburgh,  Pa. 


/ 


TRANSACTIONS    I.    E.    S. — PART    I 


13 


Sustaining  Members. 

The    following    additional    applicants 
have  been  elected  sustaining  members  of 
the  society : 
Alexalite  Company. 

432    East   23rd    Street,    New   York. 

N.Y. 
Board  of  Water  Commissioners. 

City  Hall,  London.  Ontario,  Canada. 
Cooper  Hewitt  Electric  Company. 

Eighth  and  Grand  Streets,  Hoboken, 

N.J. 
Edison   Electric  Illuminating   Co.  of 
Brockton. 

42  Main  Street,  Brockton,  Mass. 
H.  W.  Johns-Man ville  Company. 

Madison    Avenue    and    41st    Street, 

New  York. 
Little  Rock  Railway  &  Electric  Com- 
pany. 

IIS    West   4th    Street,   Little   Rock, 

Ark. 


Pittsburgh  Lamp,  Brass  &  Glass  Co. 

Pittsburgh,  Pa. 
The  Leeds  &  Northrup  Company. 

4901   Stenton   Avenue.   Philadelphia, 

Pa. 


Local  Representatives. 

To  extend  the  influence  and  work  of 
the  society  into  cities  and  territories  in 
which  there  are  not  sections  of  the 
society  the  Council  has  recently  ap- 
pointed eight  local  representatives.  The 
names  of  the  representatives  are  listed 
on  a  page  in  the  front  part  of  this  issue. 
These  representatives  will  communicate 
to  the  general  secretary  from  time  to 
time  information  concerning  local  devel- 
opments in  which  the  society  is  con- 
cerned. They  will  also  endeavor  to  pro- 
mote occasional  meetings  under  the 
joint  auspices  of  the  I.  E.  S.  and  local 
organizations. 


TRANSACTIONS 


OF  THE 


Illuminating 
Engineering  Society 


OCTOBER,    1913 


PART  II 


^ 


*r 


Papers,  Discussions  and  Reports 


[  OCTOBER,  1913  ] 

CONTENTS  --  PART  II 


Report  of  the  Committee  on  Progress 323 

Psychic  Values  of  Light,   Shade,    Form  and  Color.     By 

F.  Park  Lewis,  M.  D 357 

Neon  Tube  Lighting.     By  Georges  Claude  37  * 

The  Illuminating  Engineering  Laboratory  of  the  General 
Electric  Company.     By  S.  L.  E.  Rose 379 


y> 


REPORT  OF  THE  COMMITTEE  ON  PROGRESS.* 


To  the  Illuminating  Engineering  Society: 

During  the  past  year  the  science  of  illumination  has  probably 
made  greater  progress  than  at  any  other  similar  period  of  its 
history.  While  few  radical  changes  or  developments  have  been 
made  in  connection  with  light  sources,  improvements  have  been 
made  in  mechanical  construction  of  present  systems,  resulting  in 
increased  efficiency,  and  illumination  has  become  the  subject  of 
study  by  physicists,  oculists,  architects,  legislative  bodies  and 
others  as  never  before. 

INCANDESCENT  ELECTRIC  LAMPS. 

In  connection  with  light  sources  the  development  of  the 
tungsten  lamp  stands  out  most  prominently.  This,  in  high  candle- 
powers,  is  a  serious  competitor  of  the  arc  lamp  both  for  indoor 
and  outdoor  lighting,  and  has  largely  superceded  the  carbon  fila- 
ment lamp,  as  shown  by  the  following  table  taken  from  a  recently 
published  article,  and  giving  the  relative  number  of  lamps  sold.1 


Type 

Per  cent. 

Per  cent. 

Per  cent. 

Per  cent. 

Per  cent. 

Per  cent. 

1907 

1908 

1909 

1910 

191 1 

1912 

Carbon . . . 

93-27 

84.12 

68.98 

63.08 

52.90 

25.47 

Gem 

5-88 

8.58 

I5-07 

14.88 

19.00 

33-59 

Tantalum 

o-75 

I.78 

2.12 

3-57 

2.74 

1. 00 

Tungsten. 

0. 10 

5-52 

I3.83 
IOO.OO 

18.47 

25-30 

39-94 

Total  .... 

T  OO.OO 

100.00 

100.00 

99-94 

100.00 

It  is  extremely  probable  that  more  recent  figures  will  show  that 
the  tungsten  lamp,  aided  by  greatly  reduced  price,  its  high  effi- 
ciency and  the  fact  that  electric  companies  are  beginning  to  make 
free  renewals  of  tungsten  lamps  as  they  did  of  the  carbon  fila- 
ment, is  destined  to  make  the  carbon  filament  lamp  a  vanishing 
quantity.  In  this  connection  it  is  interesting  to  note  that  the 
United  States  government  has  issued  an  order  through  the  office 
of  the  supervising  architect  of  the  Treasury  Department  that 
carbon  and  metallized  carbon-filament  lamps  are  not  to  be  used 

*  A  report  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 
1  Lighting  Journal,  July,  1913. 


324  TRANSACTIONS   I.  E.  S. — PART   II 

in  any  of  the  government  buildings,  and  any  such  lamps  in  use 
at  the  time  of  the  receipt  of  the  order  must  be  removed  and  25- 
watt  tungsten  lamps  substituted.2 

There  has  been  marked  improvement  in  the  life  of  some  of  the 
higher  wattage  100-volt  types  and  in  compensator  and  train  light- 
ing lamps.  In  the  60,  100  and  150-watt,  100-volt  lamps  improve- 
ments in  life  have  taken  place  similar  to  those  made  last  year  in 
the  cases  of  the  250,  400  and  500-watt  lamps.  This  improvement 
in  quality  is  on  the  order  of  50  per  cent,  in  life  and  has  made 
possible  substantial  improvements  in  efficiency.  The  use  of  chem- 
icals— the  so-called  "vacuum  getter,"  in  the  bulbs  of  these  lamps 
for  reducing  the  blackening  of  globes  has  largely  brought  about 
these  improvements.3' 4  At  present  the  commercial  efficiencies  of 
these  types  are  around  1.17  w.  p.  c.  while  the  higher  wattage  types 
are  put  out  at  about  1  w.  p.  c.3> 4 

To  meet  a  demand  for  a  60-watt  small  base  lamp  for  use  in 
residence  lighting  and  to  provide  a  40-watt  lamp  that  will  in  all 
cases  be  interchangeable  with  the  25-watt  lamps  these  two  types 
are  now  made  in  smaller  bulbs.  The  60-watt  lamp  is  now  made 
in  the  same  bulb  as  was  standard  for  40-watt  lamps  and  the  40- 
watt  lamp  in  the  bulb  which  is  now  standard  for  25  watts.  This 
change  has  been  made  without  impairing  the  life  or  efficiency  of 
the  lamps.3 

Improvements  in  the  tungsten  wire-drawing  process  have  made 
it  possible  to  manufacture  wire  of  almost  exactly  the  desired  size 
and  so  render  it  possible  for  manufacturers  to  make  lamps  of  so 
nearly  the  desired  voltage  and  efficiency  that  the  function  of 
photometry  in  the  lamp  factory  is  now  principally  to  guard 
against  errors  in  manufacture.  This  improvement  has  been 
adopted  by  most  of  the  manufacturers  in  this  country  for  all 
lamps  intended  for  series  burning  service  and  has  been  productive 
of  greatly  improved  results  in  all  cases.3 

A  10- watt,  100  to  130-volt  lamp  and  a  5-watt,  50  to  60-volt 
lamp  have  been  developed.  These  are  important  additions  to 
the  line  of  sign  lamps  in  that  they  furnish  a  low  wattage  lamp 

2  Electrical  World,  March  15,  1913. 

3  Electiical Journal,  June,  1913. 

4  Electrical  Journal,  January,  1913. 


REPORT  OF  THE  COMMITTEE  ON   PROGRESS  325 

that  does  not  require  the  operation  of  more  than  two  lamps  in 
series  upon  100  or  200-volt  circuits. 

Miniature  lamps  of  the  candelabra  and  decorative  types  with 
tungsten  filaments  have  been  developed  for  no-volt  circuits. 
These  have  the  filaments  wound  to  form  a  helical  coil  of  small 
diameter  and  this  helix  is  then  mounted  on  the  supports  in  the 
same  manner  as  an  ordinary  filament.  The  mandrel  around 
which  the  filament  is  formed  is  so  small  that  the  helical  con- 
struction of  the  filament  can  scarcely  be  recognized  with  the  naked 
eye.  This  style  of  construction  renders  possible  the  use  of  fewer 
supports  and  smaller  bulbs  than  ordinary  tungsten  lamp  construc- 
tion would  allow.3 

During  the  past  year  the  use  of  concentrated  filament  tungsten 
lamps  has  enormously  increased.  At  the  present  time  a  very 
large  proportion  of  the  automobiles  now  being  sold  in  this  coun- 
try are  equipped  with  these  lamps.  Concentrated  filament  lamps 
are  also  being  used  for  stereopticon  work,  for  trolley  car  head- 
lights and  for  theatre  stage  lighting  in  place  of  arc  lamps  and 
Nernst  lamps.  With  the  style  of  construction  now  used  excellent 
concentration  is  obtained;  the  filament  for  a  ioo-watt,  115-volt 
lamp  occupies  a  space  only  7/16  inch  (11.113  mm.)  in  diameter 
and  7/16  inch  long. 

A  new  filament  consisting  of  an  alloy  of  tungsten  has  also  been 
brought  out.  It  is  claimed  that  this  not  only  possesses  the  strength 
and  efficiency  of  drawn  tungsten,  but  that  it  will  withstand  crys- 
tallization for  a  longer  period.0 

An  investigation  has  been  made  upon  the  heating  of  screw 
thread  lamp  sockets.  This  was  found  to  be  almost  entirely  due 
to  the  heat  dissipated  by  the  lamp  or  radiator  and  has  little  to  do 
with  the  energy  lost  in  contacts.  Bayonet  holders  are  satisfactory 
up  to  250  watts,  while  screw  sockets  can  transmit  energy  up  to 
1,000  watts  with  satisfaction.6 

Possibly  the  most  startling  development  in  the  tungsten  lamp 
is  the  announcement  of  one  giving  an  efficiency  of  0.5  watt  per 
candle.  A  specially  shaped  tungsten  filament  is  used  in  a  bulb 
containing  an  inert  gas,  as  nitrogen,  at  a  pressure  of  about  one 

6  Electrical  World,  December  7,  1912. 
6  Electrician,  June  27,  1913. 


326  TRANSACTIONS    I.  E.  S. — PART    II 

atmosphere.  The  types  to  be  developed  first  are  adapted  to  high 
current  consumption,  say  6  amperes  and  over.7 

A  tungsten  lamp  has  recently  been  brought  out  which  gives  a 
color  very  closely  resembling  that  of  daylight.  It  is  not,  how- 
ever, an  exact  equivalent,  and  is,  therefore,  not  suited  to  color 
matching.  The  energy  consumption  of  this  lamp  is  about  1.4 
watts  per  candle,  and  it  has  a  candle-power  of  from  70  to  75.8>  9 

A  method  of  testing  for  spots  in  incandescent  lamp  filaments 
lias  been  developed  in  which  the  body  of  a  filament  is  made  to 
disappear  against  a  luminous  background  until  the  light  or  dark 
spot  shows  by  contrast.10 

Experiments  were  made  to  determine  the  variation  in  candle- 
power  produced  by  frosting.  The  distribution  curves  were  found 
to  be  very  much  the  same  before  and  after  frosting,  and  half- 
frosting  resulted  in  a  reduction  of  from  10  to  20  per  cent,  in 
the  candle-power.  One  fact  brought  out  in  these  experiments  was 
of  practical  use  in  permitting  a  close  estimate  to  be  made  of  the 
efficiency  of  a  half  frosted  bulb.  It  was  found  that  the  curves 
for  the  distribution  of  light  of  each  lamp,  clear  and  half  frosted, 
intersected  at  an  angle  very  close  to  66°  in  each  case,  and  the 
candle-power  at  this  angle  is  in  each  case  very  nearly  45  per  cent. 
of  the  horizontal  candle-power  of  the  clear  lamp.  Thus  the 
horizontal  candle-power  of  the  clear  bulb  may  be  obtained  by 
multiplying  that  obtained  with  the  half  frosted  bulb  at  66°  by 
2.22.11 

A  focusing  tungsten  lamp  with  a  spherical  bulb  silvered  near 
the  socket  has  been  introduced  within  the  past  year.  This  lamp, 
which  is  rated  at  32  candle-power  has  an  efficiency  of  1  watt  per 
candle.  In  another  lamp  there  is  a  specially  shaped  reflector 
in  close  proximity  to  the  filament.  The  leading-in  wire  enters 
at  the  apex  of  the  reflector  and  leaves  at  its  circumference.12 

A  system  has  been  developed  in  which  14-volt  lamps  are  used 
with  a  special  transformer  for  each  lamp  or  for  each  group 
of  lamps.    Each  transformer  is  connected  in  parallel  to  the  line, 

'  Electrical  World,  July  19,  1913. 

8  Zeit.furBcleucht.,  April  30,  1913. 

9  Elek.  Anzeig.,  April  24,  1913. 

10  Electrical  World,  May  3,  1913, 

11  Electrician,  April  25,  1913. 

1-  Elek.  Anzeig.,  April  24,  1913. 


REPORT  OF  THE  COMMITTEE  ON   PROGRESS  327 

and  when  it  is  disconnected  the  primary  circuit  is  broken  so 
that  there  are  no  no-load  losses.  These  lamps  are  cheaper  than 
the  standard  no-volt  lamps,  in  addition  to  which  advantage  are 
the  long  life  and  the  possibility  of  using  low  candle-power. 
The  lamps  are  thought  to  be  especially  suitable  for  agricultural 
districts.13 

In  order  to  avoid  the  excessive  glare  inherent  in  filament 
lamps,  a  patent  has  been  taken  out  according  to  which  the  outer 
surface  of  the  lamp  bulb  is  made  in  a  series  of  fine  substantially 
parallel  grooves  constituting  a  system  of  prisms.  The  grooves 
may  be  arranged  in  various  ways.14 

ELECTRIC  ARC  LAMPS. 

The  open  and  enclosed  series  and  multiple  arc  lamps  are  fast 
disappearing,  their  places  being  taken  by  the  magnetite  and  the 
flame  arc  lamps.  There  seems  to  be  during  the  past  year,  as  far 
as  arc  lamps  are  concerned,  a  strong  tendency  towards  the  use 
of  larger  units  and  the  cutting  down  of  the  intrinsic  brilliancy 
of  the  arc  even  at  a  sacrifice  of  efficiency. 

The  flame  arc  lamp  is  of  special  value  in  lighting  large  areas, 
and  is  particularly  adapted  to  smoky  and  dusty  places,  such  as 
foundries,  blacksmith  shops,  and  railroad  train  sheds.  In  fact 
one  of  the  largest  railroads  in  the  country  spent  months  in  trying 
out  various  systems  of  lighting  for  the  train  shed  of  one  of  their 
large  stations,  and,  after  experimenting  with  various  systems  of 
lighting  finally  decided  to  use  the  flame  arc  lamp  exclusively, 
adopting  it  also  for  the  yard. 

This  lamp  has  been  improved  so  that  the  fumes  thrown  off 
by  the  arc  are  condensed  and  so  are  prevented  from  forming  a 
deposit  on  the  globe  and  from  escaping  into  the  air.  The  lamp 
is  also  economical  as  far  as  maintenance  is  concerned,  as  one 
set  of  carbons  will  burn  from  100  to  120  hours  without  attention. 

A  rather  novel  device  is  an  arrangement  for  converting  any 
enclosed  arc  lamp,  alternating  or  direct  current,  series  or  mul- 
tiple, and  regardless  of  voltage,  into  a  flame  arc  lamp.15  Many 
minor  improvements  have  been  made  in  construction  of  the  arc 

18  Electrical  World,  December  2S,  1912. 
14  Electrical  World,  September  14,  1912. 
lb  Electrical  World,  March  1,  1913. 


328  TRANSACTIONS    I.  E.  S. — PART    II 

lamp  such  as  making  the  clutch  work  on  the  electrode  indirectly, 
thereby  keeping  it  independent  of  the  size  of  the  electrode,  im- 
proving the  feeding  arrangement,  etc.  The  lamps  are  now  made 
to  give  any  desired  range  of  color  and  to  work  under  any  ordi- 
nary condition  even,  as  in  one  lamp  recently  introduced,  on  a 
25  cycle  circuit,  and  have  reached  an  efficiency  of  0.25  watt  or 
better  per  candle  with  possibilities  of  further  improvement. 

One  of  the  most  interesting  applications  of  the  long-hour 
series  alternating  current  flame  carbon  arc  lamp  is  to  low  fre- 
quency circuits  of  25  cycles.  With  the  unimpregnated  carbon 
lamp,  there  is  a  flicker  at  each  reversal  of  polarity  which,  at 
this  frequency,  is  very  marked.  In  the  flame  arc  lamp,  the  light 
is  obtained  from  the  arc  screen  and  is,  therefore,  independent 
of  polarity.16 

A  discussion  was  reported  about  the  end  of  last  year  in  which 
were  reviewed  the  results  produced  upon  the  stability  of  the 
electric  arc  by  wind  currents,  magnetic  fields,  movements  and 
presence  of  solid  obstruction,  etc.17  In  an  investigation  made 
in  regard  to  the  evaporation  temperature  of  the  arc  lamp  it  was 
found  that  the  temperature  is  independent  of  the  current,  but 
varies  with  atmospheric  pressure,  proving  that  the  atmosphere 
of  the  crater  is  at  the  boiling  temperature  of  carbon.  A  peculiar 
phenomenon  noticed  was  that  the  positive  crater  seems  to  begin 
to  boil  at  pressures  below  atmospheric.18 

VACUUM  TUBE  LAMPS. 

One  of  the  more  recent  developments  of  the  mercury  vapor 
lamp  is  a  combination  with  the  tungsten,  an  apparent  white  light 
being  thereby  produced.  The  unit  is  very  compact  and  is  fur- 
nished with  a  novel  starting  device.  The  consumption  of  this 
lamp  is  claimed  to  be  0.73  watt  per  mean  hemispherical  candle- 
power.19  Another  recent  form  of  the  mercury  vapor  lamp  is  one 
with  a  specially  designed  quartz  tube  and  designed  particularly 
for  street  lighting.  This  lamp  is  started  by  means  of  a  heated 
spiral  which  vaporizes  a  small  portion  of  the  mercury;  this  is 
automatically  cut  out  of  circuit  when  the  lamp  is  in  operation, 

16  Electrical  World,  July  26,  1913. 

17  Electrical  Journal,  December,  1912. 

18  Jnl.fur  Gasbel.,  July  12,  1913. 

19  Illuminating  Engineer  (L,ondon),  October,  1912. 


REPORT  OF  THE  COMMITTEE  ON   PROGRESS  329 

and  renders  tipping  unnecessary.  The  lamp  is  recommended  for 
the  varying  voltages  met  with  on  traction  circuits  and  is  said 
to  consume  about  0.4  watt  per  candle.20-  21 

Another  development  has  been  the  use  of  tungsten  instead  of 
platinum  wires  for  sealing  in  of  the  electrode  wires  and  the  em- 
ployment as  a  sealing  material  of  a  special  high  temperature 
glass,  having  nearly  the  same  coefficient  of  expansion  as  the 
tungsten.  Graded  glass  is  inserted  between  the  sealing  glass 
and  the  quartz  chamber  walls  in  order  to  effect  a  suitably  graded 
transition  from  the  quartz  to  the  sealing  material.22 

A  new  method  has  been  devised,  applicable  to  any  lamp  de- 
pending upon  a  vacuum,  of  sealing  the  conductors  through  the 
glass,  and  by  which  perfect  seals  may  be  obtained  between  easily 
oxidizable  metals  and  glass  of  low  fusing  point.  After  inserting 
the  conductor  through  the  aperture  in  the  glass,  the  latter  is 
strongly  heated  by  means  of  a  blow  pipe  flame  until  perfect  co- 
hesion has  been  obtained  between  the  glass  and  the  metal.  The 
seal  is  then  taken  out  of  the  flame,  and  when  it  reaches  a  dull 
red,  the  leading-in  wire  and  the  glass  surrounding  it  are  cooled 
by  several  immersions  in  a  special  bath,  say  of  sperm  or  other 
oil,  wax  or  fat.23 

The  value  of  the  ultra-violet  rays  of  the  quartz  tube  mercury 
vapor  lamp  is  becoming  recognized.  The  lamp  is  now  used  for 
sterilizing  and  the  destruction  of  bacteria.  This  effect,  however, 
decreases  in  time,  due  partly  to  the  formation  of  an  obscuring 
deposit.24 

Announcement  has  also  been  made  of  a  vapor  lamp  giving  a 
white  light.  In  this  the  tube  is  filled  with  the  vapor  of  cadmium 
with  from  3  to  to  per  cent,  of  mercury.  At  its  most  economical 
point  the  specific  consumption  is  0.16  watt  per  candle.25 

A  new  source  of  light  for  photo-electric  work  and  said  to 
have  an  effect  250  times  as  great  as  the  mercury  lamp  has  been 
introduced.  This  is  a  vacuum  discharge  tube  containing  hydrogen 
at  low  pressure.26 

20  Electt  otech.  Zetlschr.,  March  20,  1913. 

81  Elek.  Am.,  February  13,  1913. 

■  Electrical  World,  May  10,  1913. 

23  Electrician,  July  4,  1913. 

-*  Soc.  Int.  Elect,  Bull.  2,  Ser.  3,  June,  1912. 

26  Electrotech.  Zeitschr.,  Septembers,  1912. 

20  Physical  Review,  April,  1913. 


33°  TRANSACTIONS    I.  E.  S. — PART    II 

Various  investigations  have  been  made  upon  the  characteris- 
tics of  neon  lamps.  It  was  found  that  when  the  tube  was  placed 
in  an  alternating  current  4,000-volt  circuit  with  a  transformer 
and  condenser,  0.72  amperes  usefully  traversed  the  tube  Intro- 
duction into  the  circuit  of  more  neon  tubes,  instead  of  decreasing 
the  current,  apparently  increased  it  up  to  four  tubes,  after  which 
the  current  diminished.  The  phenomenon  was  not  explained, 
but  it  was  evidently  not  due  to  resonance.27 

Disappearance  of  various  gases  by  passing  a  discharge  for 
some  time  through  vacuum  tubes  was  the  subject  of  another 
investigation  and  was  found  to  be  due  to  definite  chemical  action 
rather  than  to  physical  absorption.28 

Investigation  made  also  of  the  cause  of  the  short  life  of  neon 
tube  lamps  showed  that  it  was  due  to  absorption  of  neon  by 
the  electrodes,  and  that  if  the  current  density  at  the  electrodes 
was  small,  the  life  of  the  lamps  would  be  greatly  increased.  By 
employing  very  large  electrodes,  therefore,  tubes  showed  no 
deterioration  in  2,000  hours'  use.29' 30 

GAS  LAMPS  AND  APPURTENANCES. 

A  large  number  of  new  types  of  gas  lighting  units  have  made 
their  appearance  in  both  hemispheres.  Some  of  them  have  been 
designed  with  a  view  to  more  universal  adaptability,  such  as  a 
burner  with  a  curved  bunsen  which  may  be  turned  either  up  or 
down  and  is  therefore,  applicable  to  either  upright  or  inverted 
burners  ;31  some  with  a  view  to  ease  of  maintenance,  such  as  one 
in  which  the  main  feature  is  the  possibility  of  lifting  from  the 
fixture,  without  the  use  of  pliers  ;32  and  some  with  a  view  to 
increased  efficiency,  such  as  one  in  which  the  mixture  of  gas 
and  air  issue  at  high  velocity;33  resulting  in  an  efficiency  claimed 
to  be  double  that  of  the  standard  type;  also  another  lamp  with 
an  upright  bunsen  and  an  inverted  mantle.34  Low-pressure  high- 
efficiency  lamps,  particularly  for  out-door  use  and  giving  1,500 

27  Le  Genie  Civil.  December  14,  1912. 
23  Electrician,  November  15,  1912. 

29  Electrician,  May  30,  1913. 

30  Comples  rendus,  April  28,  1913. 

31  Journal  of  Gas  Lighting,  October  1,  1912. 

32  Journal  of  Gas  Lighting,  September  24,  1912. 

33  Le  Genie  Civil,  November  23,  1912. 

34  Journal  of  Gas  Lighting,  September  14,  1912. 


REPORT  OF   THE  COMMITTEE  ON   PROGRESS  33 1 

candle-power  and  with  an  efficiency  of  nearly  40  candles  per 
foot  have  been  introduced  abroad,  but  have  not  yet  appeared  in 
this  country. 

What  is  probably  the  most  important  development  in  gas 
lighting  during  the  past  year  is  the  introduction  of  a  high-power 
single-mantle  inverted  lamp  filling  a  place  between  the  old  small 
unit  and  the  multiple  mantle  lamp.  The  lamp  gives  from  180  to 
225  mean  spherical  candles  according  to  the  gas  used  and  the 
pressure,  and  thus  occupies  a  space  hitherto  unfilled.  A  valuable 
feature  of  this  lamp  is  the  fact  that,  because  of  the  single  mantle 
the  light  is  concentrated  and  reflectors  and  glassware  may  be 
designed  with  much  greater  precision  than  with  the  multiple 
mantle  lamp.35 

Hitherto  one  of  the  difficulties  in  connection  with  the  use  of 
high-pressure  lamps  was  the  heat  developed  which  limited  the 
kinds  of  glassware  available  and  also  necessitated  a  globe  of  very 
great  size.  In  a  modification  of  one  of  the  high-pressure  inverted 
lamps  now  in  wide  use,  particularly  in  England,  a  small  silica 
cup  about  5  inches  high,  and  therefore  but  little  larger  than  the 
mantle  itself,  has  been  substituted  for  the  globe  hitherto  used. 
This  cup  totally  excludes  the  secondary  air  supply,  all  air  for 
combustion  being  admitted  as  primary  air.  The  lamp,  moreover, 
is  greatly  reduced  in  size  and  the  reflector  is  omitted.  It  is 
claimed  that  after  allowing  for  the  absorption  of  light  by  the 
silica  cup  a  10  per  cent,  increase  in  efficiency  is  obtained.  These 
cups  are  made  for  lamps  of  from  60  to  1,500  candle-power. 
They  are  very  durable  and  will  not  break  if  the  mantle  gives 
way  and  the  flame  strikes  them.36- 37 

A  new  high-pressure  lamp  has  also  been  developed  in  the 
United  States,  designed  particularly  for  street  lighting,  and 
working  under  a  pressure  of  55  inches  of  water  or  2  pounds 
per  square  inch.  These  lamps  are  at  present  made  in  two  sizes — 
500  and  1,100  candle-power  respectively  and  have  an  efficiency 
of  over  50  candles  per  foot  of  gas  per  hour.  The  mantles  are  of 
artificial  silk  and  have  a  life  of  over  400  hours  with  practically 
no  depreciation  in  candle-power.38    High  pressure  lighting,  how- 

85  Lighting  Journal   July,  1913. 

38  Journal  of Gas  Lighting,  February  18,  1913. 

3:  Journal  of  Gas  Lighting,  February  25,  1913. 

"  American  Gas  Light  Journal,  December  30,  1912. 


332  TRANSACTIONS   I.  E.  S. — PART   II 

ever,  has  not  been  developed  in  the  United  States  to  the  same 
extent  as  abroad. 

The  artificial  silk  mantle  is  steadily  coming  into  use,  its  dur- 
ability, long  life  and  strength  being  its  recommendations.  By  a 
recent  improvement  in  this  type  of  mantle,  it  can  be  burned  di- 
rectly on  the  burner,  so  that  all  burning  off,  collodionizing,  etc., 
becomes  unnecessary.39 

A  most  interesting  and  valuable  investigation  was  made  to 
determine  the  cause  of  the  falling  off  in  candle-power  of  the 
incandescent  mantle.  Microscopical  examination  showed  that 
when  new  the  ash  seemed  to  be  made  up  of  very  small  particles, 
making  an  opaque  and  light  giving  structure.  As  the  mantle  was 
heated,  these  particles  gradually  fused  together  to  form  larger 
particles  and,  by  degrees  a  large  transparent  mass,  which,  by 
the  laws  of  physics,  is  comparatively  without  light  giving  prop- 
erties.40 

Other  experiments  were  made  to  apportion  to  each  part  of 
the  lamp  its  share  of  the  deterioration.  Under  favorable  con- 
ditions, the  deterioration  in  1,000  hours'  continuous  burning  was: 

Per  cent. 

For  the  mantle  alone zyi 

For  the  burner  alone 2>£ 

For  the  glassware  alone 10 

Total  deterioration 15 

of  which,  by  proper  maintenance,  12^  per  cent,  could  have  been 
eliminated.41 

Automatic  gas  ignition  is  having  a  steady  growth.  Pilot  lights 
are  increasing  in  number,  while  the  jump-spark  and  filament 
igniters  are  gradually  coming  into  use.  Distance  control  by 
means  of  electrically  operated  gas  cocks  and  also  pneumatic  con- 
trol in  connection  with  pilot  and  jump-spark  ignition  are  being 
extensively  employed,  particularly  in  the  past  year.  It  should 
be  noted  here  that  pilot  lights  have  been  improved  by  adopting 
the  bunsen  principle  and  by  better  protection,  so  that  the  danger 
of  extinction  is  greatly  reduced.  Self-kindling  mantles  are  also 
receiving  attention,  and  one  improvement  recently  made  in  this 
type  of  mantle  was  to  mix  ammonium  chloride  with  rhodium 

33  Gas  World,  February  1,  1913. 
40  Lighting  Journal,  April,  1913. 
«  Transactions  Illuminating  Engineering  Society,  December,  1912. 


REPORT  OF  THE   COMMITTEE  ON    PROGRESS  333 

chloride.  The  metallic  rhodium  deposited  is  of  a  silver  grey 
and  comes  to  full  intensity  without  presenting  a  dark  streak  down 
the  mantle.  It  has  since  been  found  that  lithium  chloride  is 
cheaper  and  works  better.42 

The  pyrophoric  alloys  are  also  in  process  of  development,  and 
are  now  on  the  market.  A  new  composition  of  the  alloy  has  been 
found  which,  instead  of  giving  a  shower  of  sparks,  produces  a 
long  flame,  thus  rendering  the  friction  wheel  unnecessary.  This 
alloy, is  air  proof,  and  does  not  disintegrate  like  the  old  cerium 
alloy.43 

ACETYLENE. 

The  use  of  acetylene  is  increasing  for  isolated  buildings  and 
for  distribution  from  a  central  plant  in  very  small  towns.  A 
large  number  of  burners  are  available  to  suit  varying  needs,  but 
as  yet  no  incandescent  burner  has  been  devised  for  use  with 
acetylene  that  is  thoroughly  satisfactory.  An  enormous  field  for 
acetylene  has  been  developed  in  lights  for  automobiles,  where 
acetylene  dissolved  in  acetone  is  dividing  honors  with  the 
tungsten  lamp.  It  is  also  stated  that  acetylene  mixed  with  oxy- 
gen and  used  on  a  special  disk  of  ceria  gives  a  light  almost 
equal  to  the  electric  arc  and  will  shortly  be  put  to  use  with  the 
cinematograph.44 

SUNDRY  SYSTEMS. 

Air  gas  systems  and  systems  employing  kerosene  under  pres- 
sure for  use  in  connection  with  incandescent  mantles  are  mak- 
ing considerable  progress  in  the  rural  districts  and  other  places 
not  reached  by  city  gas  and  electric  plants.  The  former  system, 
however,  has  had  a  setback  in  the  increased  price  of  gasoline  and 
the  difficulty  in  obtaining  the  high  grade  necessary.  A  new 
modification  of  the  latter  system  consists  of  a  gas  arc  lamp  con- 
nected with  the  fuel  tank  by  a  fine  flexible  bronze  tube.  Labora- 
tory tests  showing  1,220  candle-power  are  claimed  for  this 
lamp.43 

A   new   method   of   utilizing   natural   gas   has    recently   been 

*-  Gas  World,  February  15,  1913. 

43  Journal  of  Gas  Lighting,  September  17,  1912. 

44  Gas  IVorld,  January  18,  1913. 
4f>  /run  Age,  January  23,  1913. 


334  TRANSACTIONS    I.  E.  S. — PART    II 

brought  out,  but  is  still  in  an  experimental  stage.  The  plan  is  to 
pump  the  nearly  exhausted  wells  and  treat  the  gas  in  such  a 
way  as  to  produce  a  volatile  liquid  which,  on  release  of  pressure 
would  evaporate  into  a  utilizable  gas. 

COLD  LIGHT. 

The  startling  announcement  has  been  made  that  the  long 
looked  for  secret  of  obtaining  a  cold  light  has  been  discovered  in 
France,  it  being  understood  that  the  term  "cold"  is  purely  rela- 
tive. A  rotating  wheel  carrying  a  number  of  lamps  at  the  cir- 
cumference is  arranged  so  that  an  electrical  contact  lights  each 
lamp  in  succession.  The  result  is  a  series  of  instantaneous 
flashes  as  each  lamp  passes  the  contact  point,  retinal  persistance 
giving  the  effect  of  a  steady  light.  With  this  apparatus,  the 
inventor  claims  to  be  able  to  run  the  lamps  at  a  voltage  of  from 
two  to  four  times  the  normal,  and  to  obtain  an  efficiency  thereby 
of  0.2  watt  per  candle,  2  watts  per  candle  being  normal.46  In 
this  connection,  it  may  be  said  that  the  idea  of  supplying  lamps 
with  intermittent  current  is  not  new.  Some  fifteen  years  ago  an 
inventor  placed  in  the  lamp  circuit  a  commutator  by  which  the 
current  was  interrupted  a  number  of  times  a  second,  but  nothing 
ever  came  of  the  plan.  A  test  of  the  present  system  is  on  record 
from  which  we  learn  that  the  results  are  what  might  be  ex- 
pected from  overrunning  the  voltage;  the  life  of  the  lamps  was 
reduced  from  one  thousand  hours  to  four  hours.47 

LUMINOUS  ANIMALS. 
The  firefly  is  still  the  subject  of  research  among  scientists.48 
It  is  pretty  well  known  that  certain  animals  and  insects  have  the 
power  to  produce  by  means  of  chemical  reactions  a  form  of 
radiant  energy  lying  practically  within  the  limits  of  the  visible 
spectrum,  but  the  nature  of  these  reactions  is  not  yet  known. 
One  investigator,  indeed,  separated  out  two  substances  called 
by  him  eluciferase  and  luciferine  respectively  which,  when  mixed 
and  moistened,  became  luminous.49  The  nature  of  these  sub- 
stances is,  however,  a  matter  of  doubt.    A  study  has  been  made 

40  Scientific  American,  May  31,  1913. 

*i  Journal  of  Gas  Lighting,  May  13,  1913. 

48  Lighting  Journal,  January,  1913. 

«  Philadelphia  Pubic  Ledger,  April  6,  1913. 


REPORT  OF  THE   COMMITTEE  ON    PROGRESS  335 

of  the  intrinsic  brilliancy  of  the  glowworm  or  larva  of  the  fire- 
fly, which  gave  as  a  result  0.0046  candle  per  square  centimeter. 
From  this  it  was  calculated  that  the  luminous  material  of  the 
glowworm,  could  it  be  reproduced,  would  not  only  be  of  high 
luminous  efficiency  but  would  also  be  a  happy  mean  in  intrinsic 
brightness,  far  lower  than  the  artificial  light  sources  with  which 
we  now  try  our  eyes,  yet  high  enough  to  permit  its  use  without 
pre-empting  more  wall  space  than  we  now  give  to  windows.50 

PHOTOMETRY. 

The  search  for  a  primary  standard  of  light  still  goes  on,  and 
the  requirements  for  one  have  been  the  subject  of  papers  ap- 
pearing from  time  to  time  during  the  past  year.  But  little  actual 
progress  in  this  direction  has,  however,  been  made  public. 

At  the  last  meeting  of  the  German  Illuminating  Engineering 
Society,  however,  a  paper  was  read  discussing  the  question  of  an 
absolute  standard,  and  it  was  proposed  to  investigate  the  dis- 
tribution of  energy  in  the  spectrum  of  the  source  of  light  to 
be  investigated,  to  determine  the  sensibility  curve  of  the  human 
eye  for  different  wave  lengths,  and  to  calculate  therefrom  the 
distribution  of  light  intensity  in  the  spectrum,  and  finally  to 
arriye  at  the  total  candle-power.  An  extended  discussion  of  the 
possibilities  of  using  the  black  body  as  the  basis  of  a  rational 
unit  of  light  followed  the  reading  of  this  paper.51 

The  past  year  has  been  noteworthy  in  regard  to  the  paucity 
of  published  reports  on  secondary  standards.  In  the  National 
Physical  Laboratory  of  England  a  set  of  twenty-four  tungsten 
substandards  running  at  1.5  watts  per  candle  has  now  been 
established  on  a  satisfactory  basis.  The  values  of  candle-power 
which  are  now  assigned  to  these  lamps  have  been  determined  by 
a  "cascade"  method  by  six  different  observers,  the  work  having 
been  in  progress  for  about  three  years.  In  the  cascade  method  it 
was  found  that  the  mechanical  errors  and  those  due  to  changes 
of  observers  from  day  to  day  have  given  rise  to  a  probable 
error  of  0.08  per  cent.  The  possession  of  a  series  of  sub- 
standards  graded  in  the  hue  of  the  light  emitted  has  been  found 
a  valuable  asset  in  the  testing  of  lamps  of  different  types  in 

60  Lighting  Journal,  February,  1913. 
51  Journal fur  Gasbel.,  July  12,  1913. 


33^  TRANSACTIONS    I.  E-  S. — PART    II 

enhancing  the  accuracy  and  facilitating  the  testing  of  such  light 
sources  as  acetylene  lamps,  gas  mantles,  flame  arcs  and  metallic 
filament  lamps  of  all  kinds.52 

Tests  were  also  made  upon  helium  tubes,  which  indicated  that 
the  maximum  deviation  in  the  tubes  tested,  and  which  were 
selected  with  care  as  to  uniformity  of  bore  and  thickness  of  wall, 
was  3  per  cent.  The  chief  improvement  now  to  be  sought  is  com- 
plete freedom  from  striations. 

In  phases  of  photometry  other  than  in  regard  to  standards,  the 
work  during  the  past  year  has  been  on  greatly  varied  lines.  Con- 
veniences in  working  the  present  type  of  photometer,  selenium 
and  other  photo-electric  cells,  the  reliability  of  the  flicker  pho- 
tometer, and  the  application  of  the  radiometer  have  all  been  the 
subject  of  study. 

The  photo-electric  cell  particularly  has  received  much  atten- 
tion. Of  great  importance  is  the  development  of  a  new  form  by 
the  pioneer  investigators  of  this  cell,  in  which  the  alkali  metal 
surface  is  covered  by  a  layer  of  hydride.  This  is  made  permanent 
by  filling  the  glass  bulb  with  a  rarefied  atmosphere  of  argon  or 
helium.    These  special  cells  possess  extremely  great  sensibility. 

Recent  experiments  undertaken  to  determine  the  relationship 
between  the  photo-electric  current  and  the  intensity  of  illumination 
have  shown  a  direct  proportionality  over  a  range  of  illumination 
from  about  one-third  that  due  to  full  sunlight  down  to  the  lowest 
illumination  detectable  by  the  eye. 

Other  experiments  have  had  for  object  the  explanation  and 
elimination  of  a  number  of  disturbing  factors  variously  called 
dark  currents  and  contact  electromotive  force  effects.  It  now 
appears  probable  that  in  the  photo-electric  cell,  when  completely 
developed,  we  shall  have  an  objective  photometer  of  very  satis- 
factory character.53 

The  selenium  cell  has  been  developed  to  such  a  point  that  it 
gives  excellent  results  with  monochromatic  light,  provided  the 
exposure  is  short.  An  accurate  sensibility  curve  has,  moreover, 
been  established  from  which  it  was  found  that  the  cell  was  most 
sensitive  to  yellowish-green  light  under  faint  illumination,  and  to 
red  light  under  strong  illumination.    The  relations  are,  however, 

62  Electrician,  June  27,  1913. 

53  Physik.  Zeilsch.,  August  15,  1913. 


REPORT  OF  THE  COMMITTEE  ON  PROGRESS  337 

so  complicated  that,  except  for  monochromatic  light  the  selenium 
cell  is  as  yet  impracticable  for  photometric  work.54 

The  efficiency  of  a  selenium  preparation  used  as  a  detector  of 
light  may  be  defined  as  the  amount  of  additional  conductivity 
imparted  to  it  by  the  unit  of  incident  light.  Later  investigations 
show  that  the  total  effect  of  the  light  action  is  shown  to  be  pro- 
portional to  the  square  root  of  the  incident  energy,  while  the 
instantaneous  effect  is  proportional  to  the  energy.  This  is  veri- 
fied down  to  an  illumination  of  o.ooooi  meter-candle.  It  is 
shown  that  selenium  is  the  most  efficient  light  detector  known, 
and  is  capable  of  discriminating  minute  differences  of  luminous 
intensity  far  beyond  the  capacity  of  the  eye.55'  56> 57 

Color  photometry  has  of  late  occupied  much  attention.  A 
chromoscope  was  devised  a  couple  of  years  ago  in  which  the 
color  could  be  determined  in  terms  of  the  colors  obtained  from 
the  standard  quartz  plates  by  polarized  light.  In  a  newer  instru- 
ment four  numbers  are  required  to  designate  the  color  under 
test  as  to  tone,  saturation  and  brightness.  The  quartz  plates  are 
in  two  systems,  but  a  wide  range  of  color  measurement  can  be 
made  using  one  system  only.  The  author  of  the  article  describ- 
ing the  instrument  considers  it  of  the  greatest  advantage  that  a 
color  can  be  reproduced  at  any  time  from  its  four  characteristic 
numbers  alone.58 

A  series  of  experiments  made  to  ascertain  the  form  of  optical 
instrument  most  suitable  for  color  measurement  showed  that  any 
color  could  be  imitated  by  the  admixture  in  suitable  quantities  of 
the  lights  of  three  suitably  chosen  narrow  regions  of  the  spec- 
trum. Instead  of  adding  these  colored  lights,  broader  primaries 
may  be  subtracted  from  the  white  light  by  the  selective  absorp- 
tion of  suitably  chosen  dyes,  but  in  this  case  the  difficulty  of 
expression  becomes  serious  owing  to  the  impossibility  of  securing 
dyes  which  will  absorb  one  primary  region  and  transmit  the 
remainder  of  the  white  light  in  equal  proportions.59 

About  a  year  ago  a  simple  method  of  comparing  the  colors  of 

64  Electrician,  December  6,  1912. 

55  Electrician,  July  11,  1913. 

66  Proceedings  Royal  Society,  August  19,  1913. 

57  Elec.  Rev.  and  West.  Elect.,  February  1,  1913. 

58  Ann,  der  Physik.,  October  15,  1912. 

59  Proc.  Phys.  Soc,  October,  1912. 


33§  TRANSACTIONS    I.  E.  S. — PART    II 

artificial  illuminants  was  described,  based  on  the  observation  of 
their  comparative  intensities  through  a  photometer.  In  the  eye- 
piece of  the  instrument  blue,  green,  yellow,  red  and  deep-red 
glasses  are  inserted  successively,  and  in  this  way  comparisons 
were  made  of  the  spectrum  intensities  of  these  colors.60 

In  connection  with  color  photometry,  the  flicker  photometer  has 
received  its  share  of  attention  and  a  discussion  has  taken  place 
as  to  whether  this  type  does  actually  measure  relative  light  inten- 
sities. It  has  already  been  pretty  well  established  that  it  gave 
more  consistent  results  when  used  to  measure  lights  of  dissimilar 
colors,  but  there  was  a  question  as  to  whether  there  was  not  a 
"luminosity  sense"  distinct  from  the  "color  sense."61  Recent 
careful  investigations  indicate,  however,  that  the  flicker  pho- 
tometer gives  true  brightness,  although  it  gives  at  low  intensities 
a  reversed  Purkinje  effect  which  makes  it  necessary  to  use  cau- 
tion in  its  employment.62 

It  has  been  suggested  that  the  physiological  process,  which 
affords  the  common  basis  by  which  colored  and  uncolored  lights 
are  measured  by  means  of  the  flicker  phenomenon,  is  the  con- 
traction of  the  iris  when  exposed  to  bright  light.  According  to 
this  the  flicker  adjustment  would  be  complete  when  the  iris  has  no 
tendency  to  alter  under  alternating  illumination.  An  experiment 
was  made  in  which  the  irises  of  the  observers  were  paralyzed 
with  a  solution  of  homatropine  sulphate.  The  appearance  of  the 
flicker  was,  under  these  conditions,  just  the  same  as  when  seen 
by  normal  eyes.  All  degrees  of  flicker  remained  as  before  the 
atropine  was  placed  in  the  eyes,  and  it  was  noted  that  the  dis- 
agreeable quality  of  the  coarse  flicker  persisted  undiminished. 
Moreover  the  numerical  results  obtained  for  each  eye  and  for 
each  observer  agree  very  well  among  themselves.63 

One  writer  claiming  to  have  a  practical  solution  of  the  problem 
of  heterochromatic  photometry  describes  absorbing  liquids,  defi- 
nitely specifiable,  which  may  be  used  in  varying  thicknesses  and 
proportions  to  make  the  light  of  a  given  standard  like  that  of  any 
other  illuminant.  A  yellow  and  a  blue  solution  have  been  found 
which  suffice  to  match  all  the  ordinary  illuminants  with  a  Carcel ; 

60  Good  Lighting,  August,  1912. 

61  Philos  Mag.,  November  and  December.  1912. 

62  Electrical  World,  April  19,  1913. 

63  Philos  Mag.,  July,  1913. 


REPORT  OF  THE   COMMITTEE  ON   PROGRESS  339 

a  purple  solution  is  suggested  for  use  where  these  are  not  suffi- 
cient, the  three  absorptions  giving  by  the  three-color  principle 
all  the  tints  as  represented  in  a  color  triangle.64 

A  suggestion  made  to  eliminate  color  difference  in  the  pho- 
tometry of  incandescent  lamps  is  to  run  the  standard  lamp  at 
such  a  voltage  that  its  color  corresponds  to  that  of  the  lamp 
under  test,  and  then  determine  the  candle-power  of  the  standard 
from  a  previously  plotted  curve  of  candle-powers  and  voltages.65 

To  the  number  of  illuminometers  already  on  the  market  has 
been  added  still  another.  This  consists  of  a  box  containing  a 
small  standardized  tungsten  lamp  at  one  end  and  a  bunsen  screen 
at  the  other.  On  one  side  of  the  box  is  a  sliding  rheostat  con- 
nected in  series  with  the  standard  lamp  and  provided  with  a 
pointer  indicating  on  a  scale  the  illumination  on  the  screen  in 
foot-candles,  the  adjustment  being  made  entirely  by  varying  the 
resistance.  Two  scales  are  provided,  one  reading  from  0.3  to 
10  foot-candles,  the  other  from  0.00 1  to  0.3  foot-candle,  cor- 
responding with  the  application  of  4  and  2  volts  respectively  to 
the  standard  lamp.66 

Still  another  recent  illumination  photometer  consists  of  a 
Lummer-Brodhun  arrangement  with  lateral  windows  to  which 
are  attached  two  rectangular  tubes,  one  directed  toward  the  screen 
upon  which  the  illumination  is  received  and  which  may  be  set  at 
any  angle  and  in  any  plane,  and  the  other  terminating  in  an 
elbow  leading  to  a  standard  tungsten  lamp,  in  which  elbow  is  set 
a  standard  reflecting  screen ;  in  this  photometer  the  standard 
tungsten  lamp  is  movable,  traveling  along  a  scale.67 

Another  photometer,  depending  upon  the  acuteness-of-vision, 
comprises  a  white  wedge  illuminated  by  the  light  to  be  tested. 
Between  the  eye  and  the  illuminated  area  is  placed  a  glass  screen 
on  which  a  diagram  of  very  fine  lines  is  marked  out.  In  com- 
paring two  lamps  the  observer  merely  moves  the  photometer  back 
and  forth  until  the  lines  become  perceptible  and  notes  the  distance 
from  the  source.  In  order  to  prevent  imagination  from  assist- 
ing the  eye,  the  lines  may  be  rotated  so  that  the  observer  does  not 
know  their  position.     The  method  is  said  to  give  satisfactory 

M  Transactions  Illuminating  Engineering  Society,  June,  1913. 

88  Electral  Review  (London),  November  15,  1912. 

•*  Illuminating  Engineer  (London),  December,  1912. 

07  Illuminating  Engineer  (London),  September,  1912. 


340  TRANSACTIONS    I.  E.  S. — PART    II 

results,  but  allowance  has  to  be  made  for  the  effect  of  adaptation 
on  the  eye.68 

During  the  year  attention  has  continued  to  be  given  to  pho- 
tometers adapted  to  the  newer  requirements  of  the  incandescent 
lamp  industry.  Two  watt-per-candle  photometers  have  been 
described  having  for  their  object  determination  of  the  voltage  at 
which  the  lamp  under  test  will  give  the  desired  efficiency  rather 
than  the  candle-power,  as  was  the  former  practice.  One  of  these 
is  noteworthy  in  that  the  operation  of  the  instrument  is  exactly 
parallel  to  the  operation  of  the  formerly  widely  used  candle- 
power  photometer  as  used  upon  single  circuit  fluctuating  volt- 
age.69 

Another  watt-per-candle  meter  has  been  devised  based  upon 
the  principle  that  if  two  potential  differences  are  equal  as  indi- 
cated by  a  galvanometer,  then  the  two  lamps  are  at  the  same 
watts-per-candle.  If  they  are  different  the  variation  in  a  resist- 
ance, required  to  make  them  equal,  is  a  measure  of  the  ratio  of 
the  watts-per-candle  of  the  two  lamps. 

A  rather  novel  method  of  reducing  the  quantity  of  light  for 
photometric  purposes  was  described  in  a  paper  read  before  the 
German  Illuminating  Engineering  Society.  If  the  light  used  for 
illuminating  the  surface  under  test  be  reduced  by  placing  in  its 
path  an  opaque  disk  with  an  open  sector,  the  illumination  of  the 
surface  will  not  be  uniform  if  the  disk  is  stationary.  The  author, 
therefore,  lets  the  light  passing  through  the  sector  be  received 
in  an  Ulbricht  sphere,  the  inner  surface  of  which  is  uniformly 
lighted  by  reflection  as  far  as  it  is  not  illuminated  by  incident 
light.  To  complete  the  arrangement  so  as  to  have  a  portable 
photometer  for  white  light,  a  white  ring  surface,  with  an  opening 
in  it  is  provided  on  the  outside  of  the  sphere.  The  observer 
changes  the  angle  of  the  sector  until  the  hole  in  the  ring  becomes 
invisible,  thus  obtaining  a  measure  of  the  light  by  which  the 
white  surface  with  the  hole  is  illuminated.70 

A  novel  method  has  been  devised  to  measure  the  energy  of  the 
ultra-violet  radiation  emitted  by  a  mercury  arc  lamp.  This  is 
based  on  the  fact  that  the  coefficient  of  velocity  of  hydrolysis  of 

68  Zeit.  Instrumenlenk.,  September  1912. 

69  Lighting  Journal,  July,  1913. 
to  Tnl.fur  Gasbel.,  July  12,  1913. 


REPORT  OF  THE  COMMITTEE  ON  PROGRESS  341 

tetrachloroplatinic  acid  varies  proportionately  to  the  incident  rad- 
iant energy.71 

Investigators  are  still  working  on  the  application  of  photog- 
raphy to  photometry  and  it  is  expected  that  an  emulsion  will 
be  obtained  which  gives  results  comparable  to  the  impression 
upon  the  eye.  Means  have  now  been  developed  for  measuring 
intensities,  opacities  and  other  properties  which  it  is  necessary 
to  know  of  the  developed  emulsion,  and  the  different  materials 
possible  to  use  in  the  emulsion  have  also  been  investigated.72'  73> 74 

An  instrument  has  been  devised  for  observing  intrinsic  bril- 
liancy. This  consists  of  a  lens  equipped  with  a  "cat's  eye" 
diaphragm  throwing  an  image  of  the  source  to  be  examined  on  a 
screen  inside  a  blackened  box.  By  means  of  a  Lummer-Brodhun 
prism,  the  observer  can  compare  the  brightness  of  the  image  with 
that  of  a  diffusing  plate  illuminated  by  a  small  standard  incan- 
descent lamp.  The  adjustment  of  equality  of  brightness  is 
made  by  means  of  the  diaphragm.75 

OPHTHALMOLOGY. 

Many  articles  have  appeared  lately  upon  the  requirements 
for  hygienic  lighting  and  the  advantages  of  one  system  of  light- 
ing over  another;  these  are  largely,  however,  either  recapitula- 
tions of  previous  knowledge  or  are  inspired  by  the  advocates  of 
the  method  of  illumination  in  question. 

An  investigation  was  made  last  year  in  regard  to  defects  of 
vision  among  school  children  in  Liverpool,  from  which  it  ap- 
peared that,  as  might  be  expected,  vision  was  better  among  child- 
ren attending  schools  in  the  outskirts  where  the  open  surround- 
ings gave  opportunity  for  open  spaces  than  among  those  attending 
schools  in  the  center  of  the  city.  With  one  exception  the  girls' 
sight  was  distinctly  worse  than  the  boys'  this  being  attributed 
by  investigation  to  the  fact  that  the  girls'  course  included  sewing. 
This  was  regarded  as  the  most  likely  of  any  item  in  the  school 
curriculum  to  affect  the  eyes,  and  is  often  carried  out  under  the 
most  disadvantageous  conditions  of  lighting.     One  investigator 

71  Comptes  rendus,  January  27,  1913. 
75  Le  Genie  Civil,  February  15,  1913. 
78  Comptes  rendus,  February  3,  1913. 
'*  Elek.  u.  masch..  May  25,  1913. 
'*  Comptes  rendus,  April  31,  1913. 


342  TRANSACTIONS    I.  E-  S. PART    II 

thinks  it  would  be  desirable  if  sewing  could  be  abolished  altogeth- 
er under  the  age  of  seven,  and  further  urges  that  there  is  not 
enough  discrimination  during  sewing  lessons  between  children 
whose  eyesight  has  been  found  satisfactory  and  those  with  ser- 
ious defects.76 

Similar  investigations  have  been  carried  on  in  other  coun- 
tries. While  it  has  been  shown  that  conditions  producing  defec- 
tive vision  are  now  much  more  closely  studied  than  heretofore, 
and  that  defective  conditions  are  being  remedied,  the  percentage 
of  children  suffering  from  poor  sight  is  still  large.  Such  sub- 
jects as  size  and  style  of  type,  quality  and  color  of  paper,  illus- 
trations, angle,  position  of  reading,  and  color  of  blackboard  and 
chalk  have  all  been,  and  are  still  being  studied,  particularly  with 
reference  to  saving  the  sight  of  the  children.77 

A  series  of  experiments  was  made  to  determine  the  perception 
of  lights  of  short  duration,  the  assistance  of  seventeen  observers 
of  different  ages  and  occupations  being  enlisted.  From  the  results 
obtained  a  curve  was  plotted  showing  the  perceptibility  of  flashes 
of  light  in  terms  of  the  durability  and  intensity.78  Tests  were 
also  made  by  other  experimenters  along  the  same  line.79 

An  interesting  investigation  was  recently  carried  on  by  a 
French  journal  to  find  the  combination  of  colors  most  legible  at 
a  distance.  The  order  of  merit  is  somewhat  surprising,  the  best 
being  black  on  yellow,  while  the  customary  combination  of  black 
on  white  appears  sixth  in  a  list  of  thirteen.80 

A  somewhat  similar  test  was  made  to  find,  if  possible,  the 
most  legible  type  to  use  for  books  and  periodicals.  The  type 
proving  best  was  the  one  selected  by  our  newest  illuminating 
engineering  journal,  while  a  rather  startling  conclusion  was 
that  the  American  typewriter  stood  in  a  class  by  itself  as  the 
worst.81 

Two  experimenters  have  recently  published  the  outlines  of 
recent  researches  lasting  over  a  year  upon  the  effects  of  radia- 
tion upon  the  eye,  most  of  the  tests  being  made  upon  rabbits.  The 

T6  Good  Lighting,  September,  1912. 

"  Ophthal.  Record,  February.  1913. 

rs  Transactions  Illuminating  Engineering  Society,  November,  1912. 

»  Electrical  World,  August  9,  1912. 

so  Scientific  American  Supplement,  February  15,  1913. 

31  Lighting  Journal ,  January,  1913. 


REPORT  OF  THE  COMMITTEE  ON   PROGRESS  343 

light  sources  chiefly  used  were  the  quartz-tube  mercury  arc  lamp 
and  the  magnetite  lamp.  Much  of  the  information  obtained  is 
not  yet  in  shape  for  publication,  but  it  was  shown  plainly  that 
under  the  ordinary  commercial  conditions  surrounding  the  use 
of  even  the  more  brilliant  illuminating  agencies,  no  specific  dan- 
ger exists ;  and  that  only  by  the  grossest  neglect  and  deliberate 
protracted  exposure  of  the  eye  to  the  brilliant  light  sources  at 
close  range  is  there  the  slightest  chance  of  injury  to  the  organs 
of  vision,  except  in  so  far  as  temporary  injury  may  be  due  to 
the  effect  of  ordinary  visible  radiation.82 

Tests  using  commercial  light  sources  and  commercial  colors 
were  also  made  to  determine  the  effect  of  a  colored  background 
upon  visual  perception.  The  results  were  not  fully  reported,  but 
they  showed  that  the  intensity  of  reflected  light  from  an  object 
is  a  much  more  important  factor  in  perceptibility  than  color.83 

Investigation  of  the  perception  of  color  shows  that  the  du- 
plicity theory  can  no  longer  be  held,  but  that  all  facts  harmonize 
with  the  theory  that  the  rods  are  the  organs  which  are  concerned 
in  the  perception  of  colors  of  short  wave  lengths.84 

It  has  frequently  been  said  that  with  natural  lighting  out  of 
doors,  a  higher  intensity  of  illumination  is  necessary  than  with 
artificial  lighting.  A  series  of  tests  were  made  at  sunrise  and 
sunset  on  a  number  of  observers,  from  which  it  was  shown 
that  daylight  illumination  in  the  open  is  satisfactory  for  reading 
at  all  intensities  as  low  as  with  artificial  light,  and  probably 
lower.85 

ILLUMINATING  ENGINEERING  SOCIETIES. 

Since  the  last  meeting  of  the  Illuminating  Engineering  Society, 
another  national  illuminating  engineering  society  has  been 
formed: — the  Deutsche  Beleuchtungstechnische  Gesellschaft. 
This  is  the  third  association  of  the  sort  in  existence,  these  beins:, 
in  order  of  time  of  formation,  the  American,  British  and  Ger- 
man. The  first  general  meeting  of  the  new  society  was  held 
in  the  physical  laboratory  of  the  University  of  Berlin  on  Feb. 
24,  1913.    Among  the  papers  presented  were:  "The  Eye  and  II- 

82  Electrical  World,  May  24,  1913. 

83  General  Electrical  Review,  April,  1913. 

84  Arch,  oj  Ophthal.,  March,  1913. 

85  Electrical  II  '01  Id,  July  5,  1913. 


344  TRANSACTIONS   I.  £.  S. — PART   II 

luminating  Engineering,"  "Light  Units,"  "The  Boiling  Tempera- 
ture of  Carbon  in  the  Arc  Lamp,"  "Method  of  Diminishing  the 
Light  Intensities  in  Photometry,"  "A  New  Method  of  Deter- 
mining the  True  Temperature  of  Filaments  in  the  Incandescent 
Lamp,"  and  "The  Present  Disadvantages  of  the  Sphere  of  Auto- 
mobile Lighting."86. 87> 88 

A  gratifying  announcement  is  that  of  the  reorganization  of 
the  International  Photometrical  Commission.  This  was  organ- 
ized by  the  International  Gas  Congress  in  1900,  and  held  its 
first  meeting  in  Zurich  in  1903.  This  was  composed  of  represen- 
tatives of  gas  companies  with  the  co-operation  of  certain  of  the 
national  laboratories.  It  is  now  proposed  to  extend  the  com- 
mission to  include  not  only  the  gas  interests  but  also  the  electrical 
interests,  and  to  be  representative  of  these  industries,  illuminating 
engineering  societies  and  other  associations  interested  in  photo- 
metry and  illumination  and  to  be  responsible  to  them.  A  sub- 
committee has  already  been  appointed  to  consider  photometrical 
units  and  standards.89'90 

The  British  Illuminating  Engineering  Society  has  partly  com- 
pleted a  study  of  school  and  library  lighting  with  a  view  to 
formulating  recommendations  suitable  for  general  adoption. 
Committees  of  this  society,  in  co-operation  with  committees  of 
the  Association  of  Teachers  in  Technical  Institutions  and  the 
Library  Association  have  issued  preliminary  reports  on  the  arti- 
ficial lighting  of  schools  and  libraries  calling  attention  to  a 
number  of  important  features  frequently  overlooked,  prominent 
among  which  is  glare.  Recommendations  as  to  minimum  illumi- 
nation for  various  purposes  were  made.91 

RELATIONS  OF  ILLUMINATING  ENGINEERNG  TO  OTHER 

BRANCHES. 

Illuminating  engineering  is  gradually  coming  into  recognition 
by  other  professions  and  arts.  The  architect  is  beginning  to 
appreciate  the  value  of  a  knowledge  of  the  principles  of  illumi- 
nation in  decoration  and  utility,  as  is  shown  by  the  co-operation 

86  Journal  of Gas  Lighting,  December  10,  1912. 

87  Electrical  World,  April  26,  1913. 

88  Elek.  Zeit.,  June  26,  1913. 

89  Journal  of  Gas  Lighting,  December  17,  1912. 

90  American  Gas  Light  Journal,  November  11,  1912. 

91  Electrician,  July  25,  1913. 


REPORT  OF  THE  COMMITTEE  ON  PROGRESS  345 

of  the  architect  and  illuminating  engineer  in  planning  the  light- 
ing features  of  the  coming  Panama-Pacific  Exposition  in  San 
Francisco,  and  in  designing  the  illumination  of  many  important 
buildings,  such  as  libraries,  churches,  restaurants,  etc.  At  the 
International  Congress  of  School  Hygiene  just  held  in  Buffalo, 
two  sessions  were  completely  turned  over  to  the  Illuminating 
Engineering  Society  and  presided  over  by  a  representative  of 
that  society.  The  theatres  are  also  coming  to  the  illuminating 
engineers  for  their  scenic  effects ;  and  in  one  installation  of 
especial  interest,  where  the  services  of  an  illuminating  engineer 
were  enlisted,  gauze  scenery  and  dyes  were  used  in  preference  to 
canvas  and  paint,  and  instead  of  using  dipped  lamps  the  effects 
were  produced  by  color  screens.  The  lighting  arrangement  is 
such  that  the  lighting  of  the  scenery  is  entirely  separate  from 
that  of  the  actors  so  that  while  the  scenery  passes  through  a 
very  large  range  of  color  changes,  the  color  of  the  light  on  the 
actors  remains  constant.  Furthermore  complementary  spots  are 
used  to  correct  the  color  of  the  actors'  faces  when  colored  light 
has  to  be  used  on  the  stage  proper.  In  this  installation  the  use 
of  the  arc  lamp  has  been  entirely  abandoned  and  the  concen- 
trated filament  form  of  tungsten  lamp  used  entirely.  Foot- 
lights have  been  largely  given  up  and  the  lighting  of  the  stage 
is  arranged  to  produce  natural  effects  by  lighting  from  overhead 
by  front  lights  in  the  gallery  and  by  lights  directed  from  one  side 
of  the  stage  so  as  to  produce  normal  shadow  effects.  It  is  in- 
teresting to  note  that  these  changes  reduced  the  maximum  load 
from  45  to  5  kw. 

Manufacturers  are  taking  up  the  question  of  lighting  their 
shops  and  mills,  and  have  investigated  the  influence  of  the 
character  of  the  light  on  the  time  taken  to  perform  mechanical 
operations.  One  of  our  large  lamp  concerns  recently  sent  a 
letter  of  inquiry  to  a  number  of  industrial  concerns;  of  the  209 
replying,  164  said  that  improvements  had  been  made  in  their 
lighting,  and  a  large  number  of  expressions  of  opinions  were 
received  to  the  effect  that  increased  .production,  better  goods, 
and  greater  satisfaction  on  the  part  of  the  workers  had  been  the 
result.  Moreover,  in  many  cases  the  cost  of  lighting  has  been 
reduced.92 

92  Journal  of  Gas  Lighting,  February  II.  1913. 


346  TRANSACTIONS    I.  E.  S. — PART    II 

Merchants  have  been  employing  the  illuminating  engineer  to 
such  an  extent  that  it  is  rare  that  a  periodical  devoted  to  the  light- 
ing industry  does  not  contain  a  description  of  the  lighting  fea- 
tures of  some  large  store  in  which  the  installation  was  carefully 
planned  and  well  carried  out.  This  applies  also  to  most  estab- 
lishments depending  for  their  success  upon  the  general  public. 

That  the  work  of  the  railway  mail  clerk  is  one  of  the  most 
trying  occupations  as  far  as  the  eyes  are  concerned  is  now  recog- 
nized, and  the  Post  Office  Department  has  issued  specifications 
for  lighting  the  mail  cars,  covering  location  of  light  source  with 
regard  to  the  initial  illuminating  values,  absorption  by  globes 
and  reflectors,  light  failures,  emergency  lighting,  etc.93  The 
data  for  these  specifications  were  obtained  from  actual  tests  on 
a  mail  car  of  one  of  our  large  railroads.94 

In  this  connection  it  may  be  well  to  mention  that  an  exhaustive 
test  has  been  made  at  the  shops  of  one  of  our  large  railway  sys- 
tems to  standardize  the  best  method  possible  of  train  lighting. 
This  test  was  made  under  the  auspices  of  the  Association  of 
Railway  Electrical  Engineers. 

The  relation  of  lighting  to  the  number  of  accidents  has  also 
been  receiving  attention  as  never  before.  The  intensity  of  the 
illumination  and  the  location  of  the  sources  have  been  studied 
in  connection  with  moving  machinery,  passage-ways  and  stairs ; 
and  in  this  connection  it  has  been  suggested  that  too  much  light 
might  be  as  bad  as  too  little  in  the  dazzling  effect  on  the  eyes. 
A  very  practical  illustration  of  the  relations  of  light  to  acci- 
dents is  the  statement  of  a  leading  casualty  company  of  New 
York  that  "the  greatest  number  of  accidents  occur  during  the 
months  of  diminished  light."  Furthermore,  a  prominent  official 
of  one  of  America's  largest  manufacturing  companies  is  author- 
ity for  the  statement  that  "insufficient  illumination"  is  frequently 
held  by  juries  to  be  "contributory  negligence,"  and  in  the  defense 
of  accident  suits  the  lawyers  of  this  company  find  it  a  valuable 
point  to  offer  testimony  by  a  competent  witness  to  prove  the 
adequacy  of  lighting  arrangements  in  this  company's  plants.95 

In  this  connection  it  may  be  said  that  several  states  have  passed 

ss  Lighting  Journal,  February,  1913.  - 
94  Electrical  World,  March  29,  1913. 
w  Safety,  p.  S6. 


REPORT  OF  THE   COMMITTEE   OX    PROGRESS  347 

laws  specifying  the  candle-power  of  locomotive  headlights,96  and 
a  number  of  tests,  the  last  under  control  of  one  of  the  state  public 
utility  commissions,  were  made  to  find  out  whether  the  lights 
complied  with  the  law  and  also  to  determine  whether  the  head- 
light might  endanger  the  safety  of  operation  of  trains  through 
interference  with  signal  lights  or  in  any  other  way.  With  elec- 
tric headlights,  numerous  phantom  lights  were  seen,  but  these 
were  of  the  nature  of  mere  flashes  where  the  engine  was  running 
at  the  ordinary  speed.  There  was  difficulty  in  distinguishing 
classification  lights  and  engine  numbers  on  locomotives  equipped 
with  arc  headlights,  and  precautions  should  be  taken  to  place  these 
marks  at  such  a  distance  from  the  headlight  that  the  latter  may 
not  materially  interfere  with  their  correct  reading.  The  arc 
lamp  should  be  switched  off  when  passing  through  large  yards 
and  other  places  where  it  might  have  a  tendency  to  interfere 
with  the  performing  of  duty  by  yardmen  or  others,  or  to  en- 
danger their  lives.  An  incandescent  lamp  should  then  be 
switched  on.97 

A  report,  rather  startling  in  its  novelty,  has  been  issued  to  the 
effect  that  souring  of  milk  is  due  to  ultra-violet  rays,  and  that 
milk  may  be  kept  sweet  for  days  by  putting  it  in  a  red  glass 
bottle,  or  in  a  plain  glass  bottle  wrapped  in  red  paper.98  A  re- 
verse action  apparently  on  this  order,  and  now  creating  much 
comment,  is  the  use  of  ultra-violet  rays  for  sterilizing  water, 
as  noted  above. 

A  suggestion  is  contained  in  an  installation  of  gas  arc  lamps 
for  lighting  tennis  courts.  Perfect  satisfaction  was  obtained,  and 
the  courts  were  even  better  patronized  at  night  when  it  was  cool 
than  in  the  heat  of  the  day.99  This  installation  paralleled  a  sim- 
ilar one  in  another  town  in  which  tungsten  lamps  were 
used,100' 101  and  these,  in  turn,  were  followed  by  a  golf  putting 
course  lighted  by  gas.  The  latest  item  of  this  sort  is  the  equip- 
ment of  a  polo-field  which  is  to  be  lighted  by  forty-eight  400- 

96  Lighting  Journal.  January,  1913. 

•7  Scientific  American  Supplement,  February  1,  1913. 

98  Scientific  American,  June  7,  1913. 

99  Good  Lighting,  September,  1912. 

100  Electrical  ll'otld,  July  12,  1913. 

101  Electrical  World,  August  24.  1913. 


34-8  TRANSACTIONS    I.  E.  S. — PART    II 

watt  tungsten  lamps  with  metal  reflectors,  the  installation  being 
designed  to  give  one  foot-candle  on  the  horizontal  plane.102 

INDIRECT  AND  SEMI-INDIRECT  LIGHTING. 

Indirect  lighting  and  its  half-brother,  semi-indirect  lighting,  are 
progressing  steadily,  and  numerous  installations  of  each  are  illus- 
trated in  the  current  technical  journals.  Manufacturers  are 
almost  daily  producing  new  forms  of  fixtures  of  this  type,  some 
of  them  artistic  in  the  extreme.  One  fixture  for  semi-indirect 
lighting  has  a  space  between  the  bowl  and  upper  reflector  en- 
closed with  clear  glass  to  keep  out  dirt,  insects,  etc.  This  has 
features  of  value,  but  its  artistic  merit  is  open  to  question. 

As  a  relief  from  the  present  standard,  though  undesirable, 
method  of  lighting  railway  cars,  it  may  be  noted  that  certain 
express  trains  now  running  out  of  New  York  are  now  provided 
with  indirect  fixtures.103 

STREET  LIGHTING. 

The  question  of  street  lighting  is  attracting  extraordinary  atten- 
tion. People  are  no  longer  satisfied  with  just  enough  light  to  see 
to  move  around  safely,  but  are  coming  to  a  realizing  sense  of  the 
advertising  and  artistic  values  of  ample  light.  White  Ways  are 
almost  as  common  as  towns  themselves,  and  where  the  cities  do 
not  seem  inclined  to  install  them,  the  merchants  put  them  in  by 
private  subscription.  The  open  arc  lamp  is  a  back  number,  and 
in  many  cases  the  enclosed  arc,  once  so  universal,  is  giving  place 
to  magnetite  or  flaming  arcs,  or  to  clusters  of  tungsten  lamps  on 
ornamental  poles.  The  old,  severely  plain  iron  or  wooden  pole 
is  rapidly  giving  way  to  the  artistic  post,  and  there  is  a  strong 
tendency  to  recognize  the  artistic  as  well  as  the  utilitarian. 

While  much  publicity  is  being  secured  to  ornamental  lighting, 
the  majority  of  the  business,  however,  has  been  connected  with 
the  ordinary  street  lighting,  which  is,  in  the  electric  field,  prac- 
tically working  in  the  direction  of  the  use  of  the  luminous  arc 
lamp,  and  the  tungsten  filament  incandescent  lamp.  The  increased 
standard  of  general  street  lighting  has  been  largely  accelerated 
by  the  reduced  cost  of  light,  and  while  it  has  taken  on  no  dis- 

102  Lighting  Journal,  May,  1913. 

103  Electrical  World,  April  26,  1913. 


REPORT  OF  THE  COMMITTEE  ON   PROGRESS  349 

tinctively  new  form,  the  standard  of  lumination  intensity  (not 
necessarily  size  of  units)  has  been  raised  very  considerably. 

An  interesting  test  was  made  in  Switzerland  to  find  the  rela- 
tive advantages  of  arc  and  metallic-filament  incandescent  electric 
lamps.  Two  streets  of  equal  length  were  lighted  with  10-ampere 
electric  lamps  and  500-candle  incandescent  lamps  respectively. 
The  choice  between  the  two  forms  of  lighting  was  left  to  29 
trolley  car  motormen.  Of  these  25  favored  the  metallic-filament 
lamp  on  account  of  lessened  glare  and  irritation  to  the  eyes.104 
The  mercury-vapor  lamp  has  been  suggested  as  a  street  lighting 
unit,  but  thus  far  its  use  for  this  purpose  has  been  hardly  more 
than  a  suggestion,  although  in  one  city  lighted  by  a  municipal 
electric  plant,  a  group  of  merchants  installed  six  quartz  tube 
lamps  to  show  by  contrast  the  poor  character  of  the  general  street 
lighting. 

During  a  "street  show"  in  one  of  our  large  cities,  ornamental 
pillars  each  carrying  three  electric  light  globes  were  erected. 
Panels  in  these  columns  were  made  transparent  by  making  them 
of  wire  netting  and  coating  them  with  varnish  of  various  colors, 
giving  the  effect,  when  lighted  from  the  inside,  of  art  glass.105 

In  the  ornamental  lighting  system,  it  is  frequently  desirable 
to  turn  off  the  lamps  without  affecting  the  rest  of  the  circuit. 
This  is  done  in  at  least  three  towns  by  various  systems  of  pilot 
wires  and  magnet  switches  centering  at  the  central  station  or 
other  convenient  point.106- 107> 108  In  this  connection  may  be  noted 
a  method  of  controlling  from  the  central  station  switches  on  a 
network  by  superimposing  a  ripple  on  the  regular  voltage.109  In 
another  town  a  switchboard  has  been  placed  in  the  office  of  the 
chief  of  police  so  that  in  case  of  burglary  or  fire  alarm  after  the 
regular  time  of  shutting  off,  the  ornamental  system  of  street  light- 
ing may  be  turned  on.110 

An  unfortunate  dispute  has  arisen  in  one  of  the  large  cities  of 
England  over  the  relative  merits  of  high  pressure  gas  and  flaming 
arc  lamps.    Two  streets  were  lighted  by  the  rival  illuminants  and 

IM  Electrical  World,  April  26,  1913. 
105  Electrical  IVorld,  October  12,  1912. 
100  Electrical  World,  November  30,  1912. 

107  Electrical  World,  September  7, 191 2. 

108  Electrical  World,  September  14,  1912. 
104  Electrician,  February  14,  1913. 

110  Electrical  World,  October  12,  1912. 


350  TRANSACTIONS    I.  £.  S. — PART    II 

experts  representing  each  of  the  industries  made  illumination 
measurements  and  prepared  reports.  The  tests  were  entirely  in 
favor  of  the  electric  lamps,  but  the  fact  was  brought  out  that  the 
gas  lamps  were  improperly  adjusted  and  installed,  so  that  no  con- 
clusions could  be  drawn  as  to  the  relative  merits  of  the  two 
systems.  The  whole  affair  caused  much  argument  and  a  good 
deal  of  acrimonious  discussion  on  the  part  of  the  advocates  of 
the  two  systems.111- 112  0 

Gas  street  lighting  has  made  great  strides  in  both  England  and 
continental  Europe,  where  high  pressure  lighting  is  in  great  favor. 
Automatic  lighting  of  gas  lamps  is  also  making  rapid  progress 
on  the  other  side  of  the  Atlantic,  the  lamps  in  a  large  number  of 
towns  being  equipped  with  these  appliances.  Two  systems  of 
automatic  lighters  are  in  extensive  use,  one  being  operated  by  a 
momentary  addition  to  the  street  main  pressure  and  the  other  by 
means  of  a  clock  arrangement  on  each  post  so  that  the  lamps 
operate  individually  and  independently.  Highly  encouraging 
reports  as  to  the  satisfactory  and  economical  working  of  these 
systems  have  been  made. 

In  this  country  street  lighting  by  gas,  while  making  steady 
progress,  has  not  experienced  the  rapid  growth  that  is  so  marked 
abroad.  Automatic  lighting  has  not  as  yet  obtained  a  foothold 
here,  and  except  for  two  minor  installations,  there  is  no  high- 
pressure  street  lighting.  Another  difference  between  European 
and  American  practise  is  that  abroad  inverted  burners  are  becom- 
ing the  universal  practise,  while  here,  except  for  a  two-mantle 
150  candle-power  lamp,  the  inverted  burner  is  not  used. 

Ornamental  street  lighting  by  gas  is  spreading,  a  number  of 
prominent  installations  having  been  made. 

It  may  be  of  interest  to  note  that  one  of  our  public  utility 
commissions,  which  had  been  investigating  the  street  lighting  in 
a  large  city  of  the  state,  recommended  that  the  city  employ  an 
illuminating  engineer,  to  be  retained  permanently  if  possible,  for 
the  purpose  of  selecting  the  type  of  lamps  to  be  used  and  to  fix 
their  location.113 

Increased  interest  in  street  lighting  does  not  seem  to  be  unusual, 

111  Journal  of  Gas  Lighting,  October  5,  1912. 

112  Electrician,  March  7  and  14,  1913. 

113  Electrical  World,  January  18,  1913. 


REPORT  OF  THE   COMMITTEE  ON   PROGRESS  35 1 

however,  as  a  recent  item  in  a  French  technical  journal  contains 
the  statement  that  in  that  country,  out  of  10,000  villages  or  com- 
munes of  more  than  1,000  inhabitants,  6,000  are  without  public 
lighting.114 

FIXTURES,  GLOBES  AND  REFLECTORS. 

Architects  and  decorators  are  realizing  more  and  more  the 
importance  of  artistic  and  appropriate  gas  and  electric  light- 
ing fixtures,  and  new  designs  adapted  to  all  conditions  are  daily 
appearing  on  the  market.  One  type  that  is  especially  popular 
just  now  is  the  "shower"  chandelier,  and  designs  of  great  beauty 
have  been  brought  out.  Another  new  design  is  an  indirect  light- 
ing portable  which  may  be  used  as  an  ornamental  table  light.115 
Still  another  fixture  designed  to  give  a  strong  concentrated  light 
for  fine  work  consists  of  a  small  reflector  socket  carrying  a 
tungsten  lamp  and  welded  to  substantial  brass  tube  bent  so  as  to 
fit  closely  the  body  of  the  machine  in  connection  with  which  it 
is  to  be  used.  This  tube  is  securely  fastened  to  the  machine 
and  becomes  practically  part  of  it.116 

A  novel  fixture  has  been  installed  in  the  rotunda  of  one  of  our 
state  buildings,  and  is  probably  the  largest  chandelier  ever  built. 
The  fixture  body  is  over  sixteen  feet  high  and  is  suspended  by 
a  chain  seventy-two  feet  long,  consisting  of  twelve-foot  links 
containing  special  tubular  tungsten  lamps  to  give  the  effect  of 
a  string  of  light,  the  joints  in  the  chain  being  provided  with  ball 
lamps  in  decorative  design.  The  fixture  itself  is  finished  in 
composition  silver  leaf.  In  some  of  the  reading  rooms  of  the 
same  building  indirect  lighting  has  been  used  with  the  bottoms 
of  the  basins  made  of  pink  Georgia  marble,  thus  producing  very 
beautiful  effects.  The  general  illumination  of  these  rooms  is 
low,  and  so  each  table  is  provided  with  reading  lights  designed 
by  actually  placing  an  individual  at  the  table  with  a  book  and 
adjusting  the  lamp  for  his  comfort. 

In  another  installation  in  an  office  building  an  attempt  was 
made  to  reproduce  artificial  daylight  through  a  false  skylight 
matching  the  color  received  through  an  actual  skylight  in  an 
adjoining  room.    The  lighting  is  so  arranged  that  at  night,  when 

114  L'Electricien,  January  n,  1913. 

1,5  Electrical  Revinu  and  (Vest.  Elect,  March  1,  1913. 

116  Electrical  World,  November  2,  1912. 


352  TRANSACTIONS  I.   E.   S. — PART  II 

artificial  light  is  used,  the  color  of  the  illumination  will  be  ad- 
justed to  match  the  artificial  lighting. 

In  connection  with  the  use  of  marble  for  diffusing  material 
noted  above,  it  is  reported  that  patents  have  been  taken  out  in 
Germany  for  using  marble  instead  of  glass.  Marble  is  planed 
down  until  it  became  translucent  and  different  intensities  of 
light  were  shown  from  behind.  The  effect  obtained  was  that  the 
illumination  was  hardly  distinguishable  from  daylight,  and  it  was 
difficult  to  realize  that  the  room  was  artificially  lighted.117 

Under  this  head  might  be  mentioned  efforts  recently  made  to 
increase  the  brilliancy  of  moving  pictures.  The  available  sources 
of  light  having  about  reached  the  limit  of  their  intensity,  the 
reflecting  power  of  the  screens  is  now  receiving  attention.  The 
early  muslin  screens  were  replaced  with  canvas  coated  with  a 
layer  of  white;  ground  glass,  which  was  next  tried,  was  found 
too  fragile,  and  was  in  turn  replaced  by  a  fabric  coated  with 
aluminum  powder  giving  a  screen  presenting  a  silvery  surface 
of  great  uniformity.  The  new  screen  is  3.7  times  as  luminous 
as  the  muslin.118 

An  investigator  conducted  a  series  of  tests  to  determine  the 
distribution  of  light  in  an  ordinary  room.  Under  the  conditions 
of  the  test  it  was  found  that  the  light  was  strongest  in  the  upper 
and  lower  part  of  the  room  and  less  intense  throughout  the 
middle  portion.  Working  on  this  idea,  a  translucent  screen  was 
made  in  the  form  of  a  vertical  half  cylinder  placed  close  to  the 
wall ;  between  it  and  the  wall  was  set  the  light  source.  Meas- 
urement of  the  light  distribution  under  these  conditions  showed 
that  it  was  almost  identical  with  that  obtained  from  daylight.119 

About  the  first  of  last  year  a  screen  was  brought  out  which' 
filtered  the  rays  of  the  electric  arc  light  in  such  a  proportion 
that  those  rays  passing  through  it  formed  a  true  daylight  color. 
This  was  followed  a  short  time  ago  by  a  similar  screen  for  the 
incandescent  gas  lamp.  The  resulting  light  is  a  perfect  match 
for  average  daylight,  and  by  it  colors  may  be  judged  with  perfect 
accuracy.120 

i"  Good  Lighting,  January,  1913. 

118  Good  Lighting,  January,  1913. 

n9  Transactions  Illuminating  Engineering  Society,  June,  1913. 

120  Lighting  Journal,  May,  1913. 


REPORT  OF  THE  COMMITTEE  ON   PROGRESS  353 

PHYSICS. 

Previous  determinations  of  the  constant  of  the  Stefan-Boltz- 
mann  law  of  radiation  vary  from  5.3  to  6.5.  During  the  past 
year,  two  independent  investigations  were  made  from  which  new 
figures  for  the  constant  were  derived.1-1- 122 

Another  investigator  describes  experiments  showing  the  devia- 
tion from  Lambert's  cosine  law  of  tungsten  and  carbon  at  glow- 
ing temperatures.  It  was  found  that  the  brightness  of  tungsten, 
beginning  with  normal  emission,  increases  with  the  angle  of 
emission,  reaches  a  maximum  at  about  750,  and  for  larger  angles 
diminished  rapidly.  The  brightness  of  carbon,  beginning  with 
normal  emission,  decreases  with  increasing  angle  of  emission, 
the  rate  of  decrease  increasing  with  the  angle.  The  relative 
brightness  variations  for  tungsten  at  the  higher  of  the  two  tem- 
peratures chosen  are  about  20  or  25  per  cent,  greater  than  the 
corresponding  variations  for  the  lower  temperature.  No  defi- 
nite change  was  found  for  carbon.123 

An  experimental  lecture  was  delivered  late  last  spring  before 
a  European  society  on  the  relations  between  spectral  analysis 
and  the  electronic  theory.  The  fundamental  point  was  that 
what  oscillates  in  light  is  nothing  but  electrons.  The  lecturer 
also  discussed  optical  resonance  as  deduced  from  the  theory 
of  electrons  and  the  Zeeman  phenomenon.124 

LEGISLATION. 

An  appreciation  of  the  importance  of  proper  illumination  is 
growing  on  the  legislative  bodies  of  different  countries,  and 
investigations  are  being  held  and  laws  passed  governing  particu- 
larly the  lighting  of  factories  and  workshops.  In  Holland,  the 
law  specifies  that  the  employment  of  women  and  young  children 
is  forbidden  in  works  in  which  artificial  light  is  normally  re- 
quired between  9  a.  m.  and  3  p.  m.  An  illumination  of  1^  foot- 
candles  is  specified  as  the  minimum  for  certain  processes  ex- 
ceptionally trying  to  the  eyes,  and  a  minimum  of  one  foot-candle 
in  less  exacting  occupations.125 

121  Electrician,  December  20,  1912. 

122  Deutsch  Phys.  Gesell.,  November  15,  1912. 

123  Astrophys.,  December,  1912. 
i24  Elek.  Zeil.,  April  23,  1913. 
154  Iron  Age,  August  17,  1912. 


354  TRANSACTIONS  I.   E.   S. — PART  II 

In  England,  a  special  committee  has  been  appointed  by  the 
Home  Secretary  to  inquire  into  conditions  necessary  for  ade- 
quate and  suitable  lighting  of  factories,  and  in  France  the  ques- 
tion has  been  under  discussion  for  about  two  years.126' 127 

In  our  own  country  the  New  York  State  Factory  Investigating 
Commission  has  been  studying  the  lighting  of  workshops  for  two 
years,128- 129> 130  and  a  bill  has  been  drafted  with  the  aid  of  a 
committee  from  this  Society  for  its  regulation.  The  Industrial 
Commission  of  Wisconsin  last  January  issued  its  general  order 
on  sanitation  which  included  ventilation  and  shop  lighting.  This 
order  provides  for  daylight  illumination  and  specifies  the  minimum 
illumination  under  different  conditions.131 

PHOTOGRAPHY  IN  ILLUMINATING  ENGINEERING. 

An  interesting  lecture  was  delivered  not  long  ago  illustrating 
the  value  of  photography  in  illuminating  engineering.  The 
speaker  laid  stress  on  the  difficulty  of  obtaining  good  photographs 
by  artificial  light,  and  said  that  there  was  little  information  avail- 
able as  to  what  the  exposure  ought  to  be  or  how  to  allow  for 
the  actinic  values  of  the  different  kinds  of  light.  The  two  essen- 
tials in  a  good  photograph  of  an  installation  are  that  the  room 
shall  appear  exactly  as  it  really  is  by  artificial  light,  and  that  the 
positions  and  natural  appearance  of  the  fixtures  shall  be  shown 
without  halation  or  distortion.  The  lecturer  believes  that  the 
ideal  way  to  preserve  a  good  record  of  lighting  installations  is 
to  have  a  really  good  photograph  showing  the  actual  installa- 
tion as  it  appeared  to  the  eye  and  also  data  on  intensity  of 
illumination.132 

In  another  lecture  on  photography  by  invisible  light,  the 
speaker  said  that  the  longest  infra-red  rays  thus  far  measured 
(those  of  the  quartz)  have  a  wave-length  of  0.3  mm.,  while  the 
shortest  electrical  waves  observed  are  2  mm.  in  length,  indicating 
a  brief  undiscovered  gap  between  the  two  sets  of  phenomena. 
The  lecturer  showed  ultra-violet  photographs   of   the   invisible 

126  Journal  of Gas  Lighting,  January  2S,  1913. 

127  Electricia?i,  January  24,  1913- 

128  Gas  World,  December  14,  1912. 

I**  Journal  of  Gas  Lighting,  December  17,  1912. 

wo  Electrical  World,  December  2S,  1912. 

131  Gas  Age,  March  15,  1913. 

is-'  Illuminating  Engineer  (London),  December,  1912. 


REPORT  OF  THE  COMMITTEE  ON   PROGRESS  355 

electronic  discharge  which  proceeds  from  the  ordinary  electric 
arc  and  not  detectible  to  the  eye.  A  current  of  air  diminishes 
the  intensity  of  this  discharge  within  its  own  range,  but  does  not 
affect  the  streamers  beyond.  The  ultra-violet  photographs  pre- 
sented were  most  interesting  in  showing  the  diffusion  of  the  ultra- 
violet shadows  even  in  full  sunshine.  Ordinary  glass  is  prac- 
tically opaque  to  light  of  this  short  wave-length,  while  the  pig- 
ment "Chinese  white"  appears  black  under  its  illumination.  The 
lecturer's  lunar  photographs  also  show  hitherto  invisible  patches 
of  heterogeneous  material  near  one  of  the  craters  indicating 
strongly  the  possibility  of  sulphur  deposits  and  so  contributing 
to  the  evidence  of  their  volcanic  origin.  The  infra-red  landscape 
views  were  remarkable  for  their  black  skies  and  strong  shadows 
and  for  the  snowy  whiteness  with  which  the  green  foliage 
appears,  owing  to  the  deep-red  component  of  its  chlorophyl  color- 
ing matter.133 

ILLUMINATION  MEASUREMENTS  AND   CALCULATIONS. 

Some  study  has  been  put  upon  methods  of  determining  and 
calculating  illumination  under  various  conditions,  and  a  number 
of  ways  of  shortening  and  simplifying  existing  means  have  been 
put  forth.  One  writer  suggests  that  if  a  curve  be  plotted  in  which 
the  y-axis,  corresponding  to  the  relation  of  the  intensities  of  the 
lights,  be  divided  to  a  logarithmic  scale,  and  the  x-axis,  corre- 
sponding to  the  distance  from  one  of  the  lights  to  the  screen,  be 
divided  to  a  natural  scale,  the  resulting  curve  will  have  a  much 
more  useful  character  than  if  drawn  on  ordinary  graph  paper.134 

A  method  has  been  worked  out  for  determining  the  illumina- 
tion at  any  point  on  a  flat  surface  illuminated  by  a  source  above 
it.  The  method  consists  in  dividing  up  the  surface  into  areas 
each  corresponding  with  a  unit  solid  angle.  By  means  of  the 
formulae  and  tables  given,  the  illumination  can  be  calculated  more 
exactly  than  by  the  approximate  methods  generally  employed.135 

One  author  criticises  the  present  method  of  laying  out  the  light 
distribution  curves  of  any  source,  and  reepmmends  an  illumina- 
tion distribution  curve  in  which  the  lengths  of  the  polar  ordinates 
are  proportional  to  the  product  of  the  intensities  into  the  areas 

138  Electrical  World,  February  8,  1913. 
184  Archiv.fur  EUklrotechnik,  1,  1912. 
la5  Eleklrolech.  Zeitsch.,  December  19,  1912. 


356  TRANSACTIONS  I.  E-   S. — PART  II 

of  the  zones  in  which  the  rays  are  taken.  The  same  writer  calls 
attention  to  the  fact  that  an  error  is  introduced  in  the  illumination 
measurements  by  neglecting  the  more  or  less  efficient  utilization  of 
the  reflected  light  according  to  the  angle  of  emission  from  the 
light  source.136 

Respectfully  submitted, 

F.  N.  Morton,  Chairman, 

F.  E.  Cady, 

E.  L.  Elliott, 

G.  L.  Hunter, 
S.  G.  Rhodes, 
Frank  E.  Wallis, 
W.  R.  Burrows, 
Dr.  F.  Park  Lewis, 
T.  J.  Litee, 
Bassett  Jones,  Jr. 

Committee. 

DISCUSSION. 

Prof.  George  A.  HoadlEy: — After  hearing  this  report  I  am 
more  than  ever  convinced  that  I  have  not  fully  appreciated  the 
progress  of  illuminating  engineering.  I  have  just  one  remark 
to  make  in  regard  to  photography  for  establishing  standards  of 
light,  and  what  I  have  to  suggest,  I  suppose  has  already  been 
done.  I  speak  of  it  simply  because  I  am  not  sure.  Those  of  us 
who  are  accustomed  to  practical  work  in  photography  know  the 
tremendous  difference  there  is  in  the  results  obtained  dependent 
upon  the  condition  of  the  illumination  of  the  subject,  and  upon 
the  length  of  time  of  exposure.  If  we  are  going  to  have  any-* 
thing  at  all  that  will  give  us  standards  along  the  line  of  actual 
illumination  through  photography,  a  standard  time  of  exposure 
and  condition  of  illumination  must  be  agreed  upon. 

136  Progressive  Age,  December  2,  1912. 


lewis:  value  of  light,  shade,  form  and  color        357 

THE   PSYCHIC   VALUES    OF   LIGHT,    SHADE,    FORM 
AND  COLOR. 


BY*  F.    PARK    LEWIS,    M.    D. 


Synopsis:  In  its  last  analysis  the  physics  of  light  must  he  considered 
in  the  effect  of  shade,  form  and  color  upon  the  human  emotions,  feelings 
and  sensations.  No  single  factor  more  definitely  dominates  the  lives  of 
men  than  the  impressions  made  upon  them  by  what  they  see.  The  dig- 
nity and  beauty  of  our  surroundings  inspire  to  higher  living  and  better 
citizenship.  We  are  brought  in  relationship  to  the  external  world  only 
through  our  special  senses.  Were  all  other  senses  abolished  while  the 
intelligence  remained,  there  would  be  no  possibility  of  communication 
with  the  outside  world.  Light  and  color  are  not  the  waves  of  different 
amplitudes  in  the  ether  but  are  the  results  of  the  impressions  produced 
upon  our  consciousness  by  these  external  influences.  We  perceive  objects 
only  by  reason  of  the  reflected  light  from  their  surfaces.  Could  we 
imagine  a  condition  in  which  surfaces  would  reflect  no  light,  objects 
could  not  be  seen  even  though  light  were  present.  The  element  of  beauty 
has  not  only  an  esthetic  value;  there  are  by-products,  if  they  may  be  so 
called,  which  are  incidentally  developed  and  which  have  even  a  greater 
bearing  upon  our  lives.  In  a  mining  district  in  Pennsylvania  the  work- 
men were  encouraged  by  competitive  prizes  to  make  beautiful  gardens  in 
back-yards  which  had  been  filled  with  debris.  The  effect  was  not  only  to 
beautify  the  district  and  to  give  an  agreeable  change  of  occupation  with 
physical  betterment,  but  an  increase  of  civic  interest.  The  mental  and 
moral  effect  of  light  and  shade  cannot  be  ignored.  Excessive  lighting 
brings  out  sordid  details  with  unpleasant  glare  which  is  as  bad  in  effect 
as  insufficient  light.  The  beauty  of  light  sources  should  not  be  forgotten. 
Good  lights  for  the  poor  will  make  the  home  beautiful.  Collaboration  is 
urged  on  the  part  of  illuminating  engineers,  architects,  school  men, 
ophthalmologists,  and  others  to  secure  broader  instruction  on  the  care  of 
the  eyes. 

It  would  seem  a  far  cry  from  the  physics  of  illumination  and 
its  application  to  the  affairs  of  civilized  life  to  a  study  in 
psychology  in  one  of  its  most  subtle  and  illusive  phases,  and  a 
word  of  apology  and  perhaps  of  explanation  may  be  necessary 
in  justification  of  my  temerity  in  asking  a  group  of  practical 
scientists  who  are  dealing  with  real  problems  and  actual  things 
to  go  so  far  afield  from  the  noise  and  bustle  of  the  street  as  to 
penetrate  into  the  mysteries  of  the  mind  itself.  I  think  you 
will  agree  with  me  however  that  the  subject  is  not  so  remote 
nor  abstract  as  might  at  first  sight  appear. 


358  TRANSACTIONS  I.  E.   S. — PART  II 

It  makes  very  little  difference  how  we  may  analyze  the  various 
spectra  in  determining  their  ultimate  construction,  or  with  what 
care  we  may  study  the  nature  of  the  light  sources  and  the  phe- 
nomena of  the  reflection  and  distribution  of  the  luminous  energy, 
if  we  fail  to  take  into  account  as  a  primary  and  essential  consid- 
eration the  effect,  not  only  upon  the  human  eyes,  but  upon  the 
human  emotions,  feelings  and  sensations,  produced  by  the  changes 
of  light  and  shadow,  of  complementary  and  contrasted  colors  and 
the  influence  actually  produced  on  our  actions  and  conduct  by 
the  things  which  constantly  come  within  the  range  of  cur  vision. 

There  is  probably  no  single  factor  which  more  definitely  dom- 
inates our  lives  than  the  impressions  made  upon  us  by  the  things 
which  we  see.  It  makes  a  vital  difference  in  his  outlook  on  life 
and  his  attitude  towards  society  if  a  man's  surroundings  are  sor- 
did and  dirty  and  mean.  Under  such  circumstances  it  will  be 
hard  for  him  to  be  a  good  citizen.  What  matters  to  him  wheth- 
er or  not  we  have  dignified  and  inspiring  architectural  monu- 
ments if  he  lives  in  a  home  but  little  better  than  a 
pig-pen?  How  can  we  hope  that  he  will  be  interested 
in  good  government,  in  public  art  galleries,  open  air 
music,  or  in  well  lighted  streets,  if  he  goes  from  digging  a 
sewer  into  a  hovel  scarcely  more  attractive?  It  is  just  the 
uplifting  glimpses  of  beauty  that  take  him  out  of  the  drudgery 
of  a  monotonous  existence  and  make  him  realize  that  his  own 
home  may  in  simple  ways  be  made  more  livable  and  more  at- 
tractive. While  this  is  in  no  sense  a  paper  on  social  welfare  the 
far  reaching  effect  of  light  and  shade  of  form  and  color  is  so 
great  as  to  excuse  the  somewhat  unusual  form  in  which  the 
subject  is  presented. 

We  know  of  course  as  a  scientific  fact  that  the  only  way  in 
which  we  are  brought  into  relationship  with  the  things  sur- 
rounding us  is  through  the  medium  of  our  special  senses,  that 
these  are  the  pathways  through  which  we  are  made  conscious 
of  the  existence  of  the  external  world.  If  one  of  these  senses 
is  lacking  or  deficient  we  must  learn  to  depend  upon  the  others. 
While  there  does  not  exist  an  automatic  or  compensatory  bal- 
ance by  which  one  sense  is  increased  in  efficiency  by  reason  of 
the  loss  of  another  the  very  necessity  of  depending  upon  those 


LEWIS  :  VALUE  OF  LIGHT,  SHADE,   FORM  AND  COLOR  359 

which  remain  to  us  may,  and  often  does,  make  them  sharper  and 
more  quickly  responsive.  An  almost  completely  blind  lad  whom 
I  saw  a  few  days  ago  was  readily  able  on  hearing  coins  jingled  in 
the  pocket  to  give  the  number  of  them  and  their  denomination. 
This  is  not  surprising  when  we  recall  that  each  piece  of  metal  has 
a  distinctive  tone  when  struck  against  another  and  that  it  required 
only  an  alert  and  correct  ear  to  differentiate  the  tones  and  to  as- 
sign to  each  its  proper  value.  The  degree  of  accuracy  possible 
when  certain  senses  are  trained  was  brought  very  vividly  to  my 
attention  sometime  since,  when  I  had  the  unusual  opportunity  of 
spending  a  few  hours  with  that  rarely  gifted  woman,  Helen  Kel- 
ler. Her  hearing  and  sight  were  lost  at  so  early  a  period  in  her 
life  as  to  exist  only  as  the  vaguest  memory.  Her  only  remaining 
avenues  of  communication  with  the  outside  world  are  through 
her  sense  of  touch,  of  taste,  and  of  smell.  The  sense  of  taste 
is  not  highly  cultivated  with  any  of  us  and  the  opportunities  in 
which  it  can  be  used  as  a  method  of  discrimination  are  relatively 
limited  and  infrequent.  There  remain  then  for  her  only  the 
sense  of  touch  which  has  been  refined  to  such  a  degree  as  to 
make  it  interpret  to  her  volumes  that  to  the  rest  of  us  are 
closed ;  and  the  sense  of  smell,  which  while  not  as  keen  as  that  of 
a  hound,  is  so  vastly  finer  than  that  which  most  of  us  possess 
that  it  serves  as  a  reliable  aid  in  a  large  number  of  circumstances 
to  enable  her  to  locate  herself,  to  determine  who  are  her  com- 
panions, and  in  a  word  to  bring  to  her  mind  a  multitude  of 
facts  for  which  most  of  us  depend  upon  the  employment  of  our 
other  senses.  The  thought  occurred  to  me  that  if  it  were  pos- 
sible to  conceive  of  the  existence  of  a  still  further  loss  of  per- 
ception so  that  there  would  be  no  method  of  conveying  to  her 
brain  the  vibrations  that  are  carried  through  the  floor  upon  which 
she  stands,  or  through  the  trunk  of  a  tree  by  which  she  recog- 
nizes the  swaying  of  the  branches  and  the  moving  of  the  leaves, 
— and  if  it  were  still  further  possible  to  imagine  the  loss  of  the 
sense  of  smell  so  that  the  odors  which  are  carried  on  the  air 
would  convey  nothing  to  her  intelligence,  and  if  with  this  were  to 
go  the  sense  of  taste  and  it  were  still  possible  that  all  her  func- 
tions could  be  carried  on,  there  could  remain  immured  within  the 
prison  walls  of  her  body  the  same  intelligent,  responsive,  percep- 


360  TRANSACTIONS  I.  t,.   S. — PART  II 

tive,  even  intutive  intelligence  that  now  exists;  but  there  would 
be  absolutely  no  method  by  which  she  could  be  brought  in  con- 
scious communication  with  the  outside  existing  world.  Indeed  so 
far  as  she  is  concerned  there  would  be  no  zvorld  because  its  exis- 
tence is  predicated  upon  its  recognition  by  the  intelligence  and 
sensation  within.  The  impressions  which  are  conveyed  to  us 
therefore  by  our  senses  are  tangible  and  real  only  to  the  degree 
that  they  are  recognized  and  understood. 

It  is  a  very  old  subject  of  discussion  as  to  whether  sound  would 
be  produced  if  a  bell  were  rung  out  at  sea  where  there  were  no 
ears  to  hear  it;  or  if  in  the  absence  of  eyes  the  trees  would  still 
be  green  and  the  poppies  red  or  the  rainbow,  when  right  condi- 
tions exist,  still  in  the  sky.  Of  course  there  is  neither  sound, 
nor  light,  nor  form,  nor  color  in  the  absence  of  an  intelligent 
recognition  of  these  qualities  because  it  is  not  the  motion  in 
the  air  produced  by  the  pounding  of  the  hammer  upon  the  bell 
which  constitutes  sound;  it  is  the  impact  of  these  atmospheric 
waves  upon  the  tympanum  carrying  an  impression  through 
sentient  nerves  to  special  brain  centers  and  the  transformation 
of  these  impressions  into  a  conscious  recognition  of  that  which 
we  understand  as  the  sound  produced  by  the  ringing  of  the 
bell.  The  colors  of  the  rainbow  are  not  the  vibrations  in  the 
ether  of  wave-lengths  of  the  different  amplitudes,  but  the  result 
of  the  impressions  which  these  vibrations  produce  upon  the 
retina,  exciting  sensations  in  the  rods  and  cones  which  are  con- 
veyed through  the  optic  nerves  to  the  visual  area  and  are  there 
interpreted  into  our  conscious  understanding  of  that  which  is 
known  to  be  the  red,  the  yellow,  the  orange,  the  blue,  the  green, 
the  indigo  and  the  violet. 

In  a  recent  discussion  before  the  Oxford  Ophthalmological 
Congress  on  "Nystagmus"  or  twitching  of  the  eyes  of  min- 
ers, which  is  produced  by  working  in  semi-darkness,  the  follow- 
ing interesting  hypothetical  experiment  was  proposed. 

Imagine  that  you  are  in  a  cavern,  the  floor,  walls  and  roof  of  which 
are  absolutely  devoid  of  color,  and  having  surfaces  of  such  a  nature  that 
they  reflect  no  light;  imagine  that  they  are  covered  with  some  substance  as 
lamp-black,  only  much  blacker. 

Put  a  lighted  candle  above  your  head,  so  that  the  light  may  fall  in 
every  direction,  but  not  into  your  eyes.  What  can  you  see?  Nothing 
but  dead  black. 


lewis:  value  of  light,  shade,  form  and  color        361 

Double  your  illumination,  have  two  candles,  and  you  will  see  "dead 
black."  Then  take  a  20  candle-power  lamp  instead  of  the  candles,  and 
you  will  see  nothing  more  than  "dead  black." 

Try  a  searchlight,  train  it  on  the  wall  opposite,  and  imagine  that  a 
large  black  beetle,  that  has  covered  itself  with  the  lamp-black-like  sub- 
stance, is  crawling  up  the  wall  in  front  of  you.  You  will  not  be  able  to 
see  it. 

It  is  impossible  to  distinguish  one  piece  of  "dead  black"  from  another. 
It  is  evident  that  under  such  conditions  increasing  the  light  does  not 
assist  vision ;  it  is  simply  wasting  light. 

Light,  in  the  absence  of  all  color,  and  in  the  absence  of  all  reflections 
from  surfaces,  is  useless  for  vision. 

Now,  imagine  that  you  are  in  a  brilliantly  colored  chamber — no  matter 
how  bright  the  colors  may  be — but  that  there  is  no  light.  What  can  you 
see?     Again,  nothing. 

Color,  therefore,  in  the  absence  of  light,  does  not  exist. 

Now  imagine  that  you  are  back  in  the  cavern  again,  which  is  devoid 
of  color  but  that  the  surfaces  are  crystalline,  such  as  coal  or  jet,  that 
is  to  say,  they  have  innumerable  small  reflecting  surfaces,  or  "surface 
brightness."  With  one  lighted  candle  above  your  head,  you  will  be  con- 
scious of  black  surfaces  with  innumerable  specks  of  glistening  white  light. 

This  is  the  light  that  is  reflected  back  directly  into  your  eyes  without 
diffusion,  and  represents  one  element  in  the  "surface  brightness"  of  the 
wall,  and  it  can  be  measured. 

Double  the  brightness  of  your  sources,  and  you  will  get  about  double 
the  amount  of  this  surface  brightness.  Have  four  times  the  light  and  you 
get  four  times  the  brightness. 

Let  the  beetle,  now  with  a  clean  and  glossy  back,  crawl  up  the  wall 
again.  You  will  be  able  to  see  a  white  spot  of  light  moving,  but  you  will 
not  be  able  to  recognize  that  it  is  a  beetle 

The  conclusions  to  be  drawn  from  this  feat  of  imagination 
are :  ( 1 )  that  light,  without  color  or  surface  brightness,  is 
useless  for  vision;  (2)  that  color,  without  light,  does  not  exist; 
(3)  that  light  without  color,  with  surface  brightness  enables 
you  to  see  a  little. 

In  this  connection  it  is  interesting  to  consider  what  are  the 
processes  by  which  we  become  conscious  of  any  existing  object. 
Let  us  imagine  what  would  take  place  in  the  brain  of  a  child 
to  whom  any  object,  an  orange,  for  example,  was  presented  for 
the  first  time. 

Every  impression  comes  as  a  new  one.  The  eyes  are  open  but 
the  images  which  have  been  carried  to  them  have  been  vague  and 
indistinct.  They  have  not  been  differentiated.  The  sound  of 
voices  comes  as  a  murmur  or  a  noise,  possibly  broken  by  varia- 


362  TRANSACTIONS  I.  E.   S. — PART  II 

tions  in  intensity,  as  from  the  crash  of  a  falling  body,  or  the 
jangling  of  a  bell;  but  the  nicer  discriminations  of  sound,  which 
may  be  developed  with  increasing  refinement  until  the  most  ex- 
act harmonies  or  the  least  discordant  note  is  at  once  recognized, 
have  not  yet  begun.  The  first  impression  conveyed  to  the  eyes 
of  the  child  after  making  the  distinction  between  light  and  dark- 
ness would  be  that  of  color.  When  the  orange  was  brought  in 
his  range  of  sight,  its  brilliancy  would  attract  his  notice  and  he 
would  become  conscious  of  a  blotch  of  color,  like  the  sunshine 
or  the  lamp-light  which  he  has  already  seen,  and  this,  in  the 
sight  center  situated  at  the  back  of  his  brain,  would  occasion  a 
flood  of  nervous  energy,  excited  by  the  vibrations  in  the  ether, 
and  the  neurons  of  the  terminal  nerve  endings  would  respond 
with  a  quickened  capacity  to  apprehend,  to  appreciate  this  same 
phenomenon  when  it  again  occurred.  It  would  be  the  beginning 
of  those  finer  color  distinctions  that  zvere  to  come  later  and 
which  were  to  constitute  the  education  of  his  color  sense,  which 
zvere  to  enable  him  to  understand,  to  appreciate  and  to  feel  the 
beauties  of  the  color  harmonies  of  the  world  in  which  he  lives. 
Then  gradually  this  splotch  of  color,  otherwise  so  meaningless, 
would  take  on  form,  and  he  would  realize  that  it  was  limited  by 
a  circle,  and  in  gaining  this  knowledge  a  new  group  of  cells 
would  be  energized,  and  another  essential  fact  in  relation  to  his 
surroundings  would  have  been  achieved.  Then  would  come  to 
him  the  realization  of  a  new  and  more  wonderful  phenomenon: 
the  circle  has  depth;  and  here  an  enormous  advance  has  been 
made.  He  has  been  introduced  into  three  dimensional  space. 
Both  sides  of  his  brain  are  working  synchronously.  He  has 
binocular  vision.  The  image  which  has  been  made  upon  the 
retina  of  each  eye  has  been  carried  to  corresponding  parts  of 
the  brain,  overlapping  and  blending  as  in  a  stereoscope.  A  mul- 
titude of  new  impressions  have  been  aroused,  suggestions  of  the 
outer  world,  indeed  of  the  universe,  have  been  conveyed  to  him. 
His  logical  faculties  have  been  awakened,  and  without  either 
knowing  or  realizing  it  he  has  done  the  most  important  thing 
in  the  world.  He  has  begun  to  think.  The  thought  which  he  is 
unable  to  express  has  aroused  his  will,  has  excited  his  desire, 
has  tempted  him  to  experiment.  He  timidly  and  tentatively 
reaches  out  his  hand  and  touches  the  thins:  that  he  has  seen  and 


LEWIS  :  VALUE  OF  LIGHT,  SHADE,   FORM  AND  COLOR  363 

in  recognizing  the  consciousness  of  its  presence  he  has  done  the 
most  wonderful  thing  in  the  world.  He  has  established  the  exis- 
tence of  a  problem,  the  solution  of  which  has  been  the  basis  of  the 
speculations  or  our  most  profound  philosophers,  from  Plato  to 
Kant.  He  has  located  himself  in  space.  He  has  begun  to  find  him- 
self. He  has  commenced  his  education.  He  will  now  learn  to  dif- 
ferentiate between  soft  and  hard,  between  rough  and 
smooth,  between  elevation  and  depression,  between  those 
things  which  oppose  and  those  which  attract.  With  each 
new  idea  has  come  a  new  flood  of  energy,  sweeping  through 
his  brain  and  making  the  pathway  easier  for  those  which 
are  to  follow.  The  skin  of  the  orange  is  broken  and 
the  fragrance  of  the  volatile  oil  is  carried  to  his  nose.  It 
brings  an  odor  like  nothing  else  in  the  world,  yet  it  is  one  of 
hundreds  of  perfumes  and  scents  and  smells  that  he  is  to  learn 
to  differentiate  from  all  others,  and  with  sensation  comes  thought, 
with  thought,  suggestion,  and  with  suggestion  will,  and  the  motor 
influences  which  are  to  govern  this  will  during  all  of  his  life  have 
been  established. 

The  orange  drops  from  the  baby  hands  and  falls  with  a  dull 
thump  to  the  floor.  It  gives  a  sound  that  to  the  trained  ear  con- 
veys intelligence  of  the  nature  of  the  thing  itself.  It  is  at  once 
evident  that  that  which  has  fallen  is  not  metallic;  it  has  not 
the  flat  sound  of  a  closed  book  nor  the  overtones  of  a  hollow 
wooden  box.  It  is  one  of  a  thousand  possible  sounds  and  yet 
it  carries  a  descriptive  story  to  the  listening  ear  and  trained 
brain.  Finally  the  fruit  is  retrieved,  the  skin  is  removed,  the 
segments  are  broken  apart,  and  everyone  of  these  movements 
with  the  little  sounds  connected  with  them  are  educative.  When 
at  last  a  portion  of  the  fruit  is  conveyed  to  the  mouth  they 
arouse, — who  can  say  how  many  groups  of  motor  influences : 
the  whole  body  is  moving,  the  neck  bends,  the  arms,  the  wrist, 
the  fingers,  each  with  its  separate  centers  in  the  brain  represented 
by  neurons,  almost  without  number.  When  the  segment  is 
placed  in  the  mouth  the  muscles  of  the  lips,  the  tongue,  the  jaws, 
the  cheeks — all  are  dominated  by  impulses  which  are  sent  out 
from  the  brain — and  finally  when  the  juice  of  the  fruit  is  tasted 
and  this  wonderful  complex  of  sensations  has  been  united  into 
what  has  been  termed  a  sterognostic  comprehension  of  the  whole, 


364  TRANSACTIONS  I.  E.   S. — PART  II 

the  entire  brain  has  been  excited  into  activity  from  front  to 
back  and  from  side  to  side  giving  instructions  in  co-ordination, 
in  will,  in  logic,  and  perhaps  in  ethics.  Each  group  of  nerve 
centers  that  has  been  energized  in  receiving  impressions  or  in 
sending  out  commands  is  being  educated  in  the  only  way  in 
which  it  can  be  to  perform  the  work  which  it  is  ultimately  des- 
tined to  do. 

So  that  not  only  the  perception  of  form  or  color  is  an  essen- 
tially psychic  process  but  by  the  automatic  relationships  which 
are  aroused  through  what  are  termed  the  association  fibers  in 
the  brain,  other  emotions  are  excited,  and  what  would  seem  to 
be  a  simple,  becomes  a  most  complicated  process.  The  thing 
that  we.  see  may  give  rise  to  emotions  far  removed  from  that 
which  might  be  naturally  expected.  For  example,  a  landscape 
of  most  extraordinary  beauty  may  be  associated  with  some 
earlier  circumstance  of  a  repellent  character  and  not  only  will 
that  particular  view  be  ever  afterward  disagreeable  but  things 
associated  with  that  view  of  which  we  may  be  quite  unconscious 
when  they  appear  in  other  places  and  under  other  conditions 
may  excite  that  same  unpleasing  sensation.  It  shows  conse- 
quently that  there  can  be  no  real  or  exact  psychic  values,  for  not 
only  are  the  things  which  we  see  modified  in  our  recognition  of 
them  by  things  which  we  have  seen,  frequently  very  early  in 
life,  but  they  may  be  so  modified  also  by  associated  contigencies 
that  there  can  be  no  exact  and  invariable  value  which  is  not 
modified  by  previous  impressions.  In  consequence  of  this  fact 
every  view  which  meets  the  eyes  is  a  composite  of  the  thing  seen 
with  an  additional  modifying  element  which  may  be  in  some 
cases  absolutely  changed  by  the  supplemental  impressions  which 
we  bring  to  it.  This  has  been  so  definitely  recognized  by  the 
modern  school  of  artists  that  they  have  long  since  realized  that 
the  more  exact  the  reproduction  of  a  landscape  or  of  a  face  the 
less  like  is  it  to  the  original.  It  is  impossible  for  any  artist 
to  place  in  his  picture  the  fugitive  impressions  which  are  rapidly 
chasing  each  other  over  the  features,  or  the  flash  in  the  eye 
which  so  illuminates  the  character,  or  the  droop  of  the  lip  which 
may  betray  a  weakness,  and  all  of  which  are  instantaneous. 
Monet  painted  seventeen  views  of  a  haystack  because 
as    the    day    changed,    it    was    never     twice    alike.     But     if 


LEWIS :  VALUE  OF  UGHT,  SHADE,  FORM  AND  COLOR  365 

he  can  suggest  in  vaguest  outline  that  which  he  feels  and  sees 
he  has  made  it  possible  for  the  understanding  observer  to  supple- 
ment in  his  own  mind  that  which  has  been  suggested.  The 
indefinite,  hazy,  shadowy,  landscape  then  becomes  vitalized  by 
the  associated  memories  which  it  has  stimulated.  With  the 
definite,  sharp-cut,  precise  reproduction  of  the  actual  form  of 
the  thing  itself  staring  him  in  the  face,  as  in  a  photographic 
representation  of  it,  all  of  these  subtler  but  therefore  more  real 
qualities  which  differentiate  that  thing  from  all  other  things  in 
the  world  are  masked.  A  mask  not  only  conceals,  it  distracts 
the  mind  from  that  which  is  beneath  it  and  prevents  that  play 
of  the  mind  which  enables  one  to  build  out  of  the  shadowy  sem- 
blance all  of  those  beauties  with  which  we  would  wish  to  see  it 
invested. 

All  men  of  imagination,  whether  they  are  artists  in  stone  or 
in  words,  whether  they  are  the  discoverers  in  science,  the  lead- 
ers in  finance  or  the  makers  of  an  empire,  are  essentially  and 
potentially  poets  and  the  poet  is  he  who  short-circuits  truth 
through  the   fourth   dimension  of   intuition. 

We  are  all  influenced  emotionally  by  form  and  color.  We 
unconsciously  feel  the  depression  produced  by  the  black  gar- 
ments of  widowhood,  and  the  enlivening  cheering  effect  of. 
brightness  and  color.  We  feel  the  color  atmosphere  in  which  we 
live.  They  who  live  in  a  land  of  clouds  reflect  in  their  charac- 
ters and  bearing  the  shadows  that  fall  upon  them ;  while  the 
sunshine  glows  in  the  lives,  in  the  mentality,  and  in  the  activi- 
ties, of  those  of  southern  lands.  As  the  heavier  and  more  mas- 
sive forms  of  architecture  have  a  morally  depressing  effect, 
as  dark  walls  and  sombre  furniture  drink  in  the  light,  as  dim 
rooms  and  badly  illuminated  corridors  are  forbidding  and  fore- 
boding, so  is  the  converse  true.  So,  too,  the  dazzling  glare  of 
brilliant  lights  brings  out  in  actual  detail  all  the  sordid  fittings  of 
a  poor  room  or  the  inharmonious  settings  of  a  badly  furnished 
one.  We  feel  instantly  the  atmosphere  of  the  place  we  enter. 
The  importance  of  beauty  as  an  essential  element  in  civic  better- 
ment and  in  its  social  aspects  has  not  received  the  attention 
which  it  deserves.  If  we  are  influenced  to  the  degree  which  I 
have  indicated  by  our  environment  it  must  follow  invariably 
that  surroundings  which  give  an  atmosphere  of  quiet  and  which 


366  TRANSACTIONS  I.  E.   S. — PART  II 

pleasurably  excite  the  imagination  must  have  a  beneficial  moral 
effect  upon  the  community.  The  atmosphere  which  exists  to- 
day is  one  of  excitement  and  the  tendency  is  to  increase  this 
excitement  by  all  forms  of  artificial  stimulant.  Our  newspapers 
not  satisfied  with  four-inch  headlines  in  order  to  attract 
notice  have  adopted  green  and  pink  outer  covers  as  an  added  ap- 
peal to  the  eye. 

THE  UTILITY  OF  ART. 

The  element  of  beauty  in  our  common  life  has  not  only  an 
esthetic  value  which  in  itself  is  important  but  there  are  certain 
by-products,  if  they  might  be  so-called,  which  are  incidentally 
developed  and  have  even  a  greater  bearing  upon  our  lives. 

In  a  mining  district  within  an  hour's  journey  from  Pittsburgh 
a  few  years  ago  the  homes  of  the  miners  were  sordid  and  un- 
kempt, the  streets  being  littered  and  the  back  yards  repositories  for 
all  sorts  of  rubbish.  The  corporation  in  charge  of  the  community 
introduced  in  the  management  of  the  plant  improvements  in  ac- 
cordance with  the  most  advanced  methods,  and  among  other 
measures  prizes  were  offered  for  the  best  garden  plots  to  be 
found  about  the  homes  of  the  workmen.  This  has  resulted  in  an 
eager  and  intelligent  competition  and  in  place  of  an  offense  to 
the  eye  the  town  has  become  a  beauty  spot.  Fences  are  lined 
with  rows  of  hollyhocks  and  golden  rod,  the  walks  are  bordered 
by  attractive  arrangements  of  garden  flowers  beautifully  kept; 
vegetable  gardens  have  become  productive  and  in  some  instances 
have  added  to  the  annual  income  of  the  miners  as  much  as  $100. 
Miners  working  in  dark  coal  shafts  straining  their  eyes 
to  see  and  having  them  dazzled  by  millions  of  reflections  from 
the  shiny  surfaces  suffer  from  an  affection  known  as  miners 
nystagmus,  the  rapid  oscillation  of  the  eyes,  to  which  reference 
has  already  been  made.  I  am  told  that  the  refreshing  change 
from  the  gloom  of  the  mine  to  the  soft  colors  of  the 
garden  in  which  they  work  has  already  exercised  a  most  bene- 
ficial influence  upon  this  serious  affection.  In  its  moral  effect  in 
giving  an  appreciation  of  the  values  of  better  civic  conditions  in 
developing  a  civic  spirit,  interest  has  been  aroused ;  they  have  not 
only  made  gardens  they  have  made  men  and  citizens. 

The  mental  and  moral  effect  of  light  and  shadow,  the  difference 


LEWIS  :  VALUE  OF  LIGHT,  SHADE,   FORM  AND  COLOR  367 

produced  upon  our  state  of  mind  by  the  glaring  brilliancy  of  an 
unshaded  Welsbach  light,  especially  an  old  one,  or  the  soft  glow 
of  an  even  yellow  illumination  is  felt  by  every  one  although  by 
no  means  always  recognized  as  a  cause  of  nervous  irritation.  In 
some  of  the  most  persistent  cases  of  eyestrain  after  the  ophthal- 
mologist has  employed  the  highest  degree  of  skill  in  determining 
the  correct  combination  of  lenses  to  be  employed  it  will  be  found 
that  the  discomfort  is  due  to  a  badly  placed  lamp,  to  the  improper 
use  or  absence  of  shades,  to  an  insufficiency  or  an  excess  of  light, 
to  some  specular  reflection,  or  other  local  fault  in  the  illumina- 
tion about  which  he  has  not  been  advised. 

There  is  probably  no  one  simple  element  that  more  deeply  con- 
cerns the  welfare  of  all  people  than  correct  lighting. 

In  the  studies  in  efficiency  in  lighting  we  seem  to  have  forgotten 
the  beauty  of  light  itself.  No  mere  luminosity  will  replace  a 
visible  light  source.  By  grouping  lights  of  low  power  and  prop- 
erly planning  them — using  translucent,  frosted,  or  prismatic 
globes,  effects  of  great  beauty  may  be  secured.  It  would  seem 
unnecessary  to  put  emphasis  too  strongly  on  the  superiority  of 
indirect  illumination. 

It  is  in  the  dark  streets  and  unlighted  alleys  that  crime  skulks. 
Good  lights  are  cheaper  than  policemen  and  more  effective. 
When  the  dark  corners  are  lighted  we  are  ashamed  to  have 
them  cumbered  with  debris  and  we  clean  them  up. 

When  a  man  comes  home  after  a  hard  day's  work,  stumbles 
through  a  dark  hallway  and  finds  the  living  room  dimly  lighted 
by  an  unshaded  lamp  with  a  smoky  chimney,  glaring,  yet  in- 
sufficient, bringing  out  all  of  the  misery  without  softening  any 
of  its  harsher  outlines,  is  it  to  be  wondered  that  he  makes  his 
stay  as  short  as  possible,  seeking  in  preference  the  brightly  lighted 
saloon  where  appeal  is  made  to  his  eyes  as  well  as  to  his  appetite  ? 
Until  we  make  the  homes  of  the  poor  fit  habitations  for  them  to 
live  in  we  cannot  expect  them  to  spend  much  time  in  them,  nor 
can  we  expect  to  make  good  citizens  out  of  them.  How  can  we 
hope  for  civic  pride  or  civic  righteousness  to  come  out  of  an 
unlovely  dirty  tenement  house.  A  stream  will  not  rise  higher 
than  its  source,  and  the  source  is  the  home  and  the  family. 

One  of  the  easiest,  one  of  the  least  expensive  methods  of  mak- 


368  TRANSACTIONS  I.  E.   S. — PART  II 

ing  the  poor  home  livable  would  be  to  introduce  good  lights  in  it. 
Could  a  more  effective  or  a  more  helpful  propaganda  be  inaug- 
urated than  to  teach  the  poor  to  light  their  homes  adequately, 
beautifully  and  cheaply?  This  could  be  done  with  a  minimum  of 
effort,  for  many  of  the  homes  are  lighted  imperfectly,  in  an  ugly 
way  and  at  an  extravagant  cost.  It  should  be  one  of  the  first  steps 
in  the  new  movement  for  the  conservation  of  vision. 

It  ought  not  to  be  a  difficult  matter  to  secure  proper  lighting 
for  our  public  institutions.  Of  all  buildings  those  which  should 
demand  good  lighting  are  our  public  libraries.  Still  to-day  when 
so  much  has  been  accomplished  on  these  lines,  it  is  the  exception 
rather  than  the  rule  to  find  a  public  library  in  which  the  lighting 
is  not  atrocious. 

IMPORTANCE  OF  GOOD  LIGHTING. 

The  importance  of  good  lighting  in  public  buildings  is  so  self- 
evident  that  it  would  not  require  mention  were  it  not  for  the  fact 
that  often  in  the  finest  specimens  of  modern  architecture  this 
seems  to  have  been  overlooked. 

Some  time  since  I  happened  to  be  in  one  of  the  progressive 
western  cities  where  the  State  house,  a  splendidly  located  edifice 
on  a  hill,  which  was  notable  for  its  beautiful  approaches,  was  so 
poorly  lighted  that  on  entering  the  relatively  small  doorway  on  a 
clear  sunshiny  morning  it  was  found  that  the  entire  main  floor 
was  artificially  illuminated  and  the  basement  floor  in  which  were 
situated  some  of  the  most  important  offices  for  the  transaction 
of  the  business  of  the  state  could  not  have  been  used  were  it  not 
for  the  artificial  lights  employed.  Unhappily  this  is  not  an  ex- 
ceptional circumstance,  alike  serious  defect  is  found  in  the  multi- 
million  dollar  State  Capitol  at  Albany,  N.  Y. 

In  our  auditoriums  the  lights,  to  paraphrase  the  meaning  of 
the  apt  French  expression,  "jump  to  the  eye."  In  unnumbered 
public  schools  to-day,  in  which  artificial  lights  must  be  used,  the 
children  are  facing  flickering  gas  lights  in  a  vain  attempt  to  see 
the  marks  on  shiny  black-boards.  The  school  authorities  have 
not  yet  learned  that  dark  red  and  green  walls  absorb  the  light  for 
which  the  children  are  suffering. 

The  time  has  fully  arrived  when  an  authoritative  body  com- 
posed of  architects,  of  illuminating  engineers,  of  school-men,  of 


VALUE  OF  LIGHT,  SHADE,  FORM  AND  COLOR  369 

ophthalmologists  and  of  all  others  who  have  to  do  with  the  man- 
agement of  light  or  the  use  of  the  eyes  should  collaborate  in  the 
development  of  authoritative  plans  for  the  education  of  the  public 
on  sight  protection.  The  American  Medical  Association  is  now 
forming  a  sub-committee  from  the  medical  societies  in  every  state 
in  the  Union  on  the  conservation  of  vision.  The  National  Edu- 
cation Associaton  is  deeply  interested  and  is  now  ready  to  sup- 
port any  proper  effort  for  broader  instruction  on  the  care  of  the 
eyes. 

It  was  proposed  several  years  ago  that  there  might  be  one 
day  in  the  year  given  to  the  conservation  of  sight.  It  would  be 
of  great  interest  for  the  children  to  study  the  condition  of  their 
own  schools.  In  this  respect  their  essays  might  include  the 
physics  of  light,  illumination,  natural  and  artificial,  the  amount 
of  window  space  necessary  for  a  well  lighted  room,  how  it  should 
be  placed,  in  a  word  the  hygiene  of  the  eyes.  The  study  which 
this  would  necessitate  would  give  a  groundwork  of  knowledge 
which  would  promote  better  conditions  in  the  future  than  exist  at 
present.  It  would  constitute  a  practical  lesson  on  one  of  the  es- 
sentials of  right  living  and  would  result  in  collateral  benefits  of 
inestimable  importance. 

DISCUSSION. 

Mr.  G.  H.  Stickney  :  I  can  not  pretend  to  discuss  this  paper. 
On  the  other  hand,  I  feel  that  we  must  express  a  special  appre- 
ciation for  it.  It  seems  to  me  that  it  is  one  of  the  most  valuable 
papers  that  I  have  listened  to  in  a  very  long  time.  To  those  of 
us  especially  who  work  largely  from  the  engineering  end  it  brings 
a  point  of  view  which  should  modify  our  thought,  and  round  out 
our  practise. 

Dr.  H.  E.  Ives:  Mr.  President,  I  wish  to  second,  if  I  may  so 
put  it,  the  remarks  of  Mr.  Stickney.  I  think  that  it  would  be 
safe  to  say  that  we  have  never  in  our  history  had  presented  to 
us  so  clearly  and  impressively  the  importance,  the  almost  sub- 
lime importance,  of  illuminating  engineering.  We  have  had 
pointed  out  to  us  among  other  things  the  importance  of  associa- 
tion. We  are  accustomed  to  a  certain  kind  of  lighting  and  are 
apt  to  argue  because  we  have  been  adapting  to  that  kind  of 


370  TRANSACTIONS  I.  £.   S. — PART  II 

lighting  through  the  ages  that  it  is  necessarily  best.  But  after 
all  much  of  it  may  be  a  question  of  association.  I  talked  not  long 
ago  with  a  prominent  member  of  this  Society,  who  told  me  that 
the  people  of  his  house  had  become  entirely  adapted  to  another 
system  of  illumination  than  the  one  to  which  they  were  formerly 
accustomed  although  the  new  system  at  first  seemed  all  wrong. 
It  opened  my  eyes  to  the  fact  that  if  the  associations  are  properly 
planned  it  may  be  that  we  can  very  easily  improve  upon  daylight, 
or  anything  else  that  you  want  to  present  as  ideal.  It  gives  us 
an  opportunity.  For  instance,  we  can  light  a  room  by  light 
from  the  side  or  light  from  overhead,  and  we  can  adapt  our- 
selves to  either  one.  We  can  allow  ourselves  to  be  guided  by 
other  considerations  than  first  impressions.  We  can  search  out 
new  lighting  effects. 

Dr.  Lewis'  has  spoken  of  the  healthful  effect  of  proper  associa- 
tions and  proper  lighting  conditions.  Has  it  ever  occurred  to 
you  that  we  might  start  a  new  school  of  medicine.  We  have  all 
sorts  of  "paths"  who  have  been  in  a  correspondence  school  for 
six  weeks  and  learned  all  there  is  to  be  learned  about  healing. 
May  we  not  look  forward  to  the  "photopath"  who  will  put  a 
patient  in  a  calm,  restful  room,  and  subject  him  to  lighting  ef- 
fects to  subdue  or  stimulate  him  until  his  nervous  condition  im- 
proves ? 

Another  point  which  Dr.  Lewis  has  made  which  I  think  is 
of  very  material  commercial  importance  is  that  the  matter  of 
lighting  is  getting  to  be  recognized  as  an  aid  to  architecture. 
Formerly  an  architect  designed  his  building  as  it  would  appear 
in  daylight.  Then  the  lighting  was  put  in  as  a  necessary  evil — 
and  it  usually  was  an  evil.  Now,  due  to  the  application  of  a  great 
many  minds — and  very  artistic  minds — to  the  lighting  problem, 
we  see  to-day  lighting  installations  which  are  appropriate.  So 
that,  speaking  for  myself,  I  would  prefer  very  often  to  see  the 
room  lighted  up  by  night  rather  than  to  see  it  by  day,  simply 
from  the  beauty  of  the  light  sources. 

As  I  said  before,  we  ought  to  be  guided  by  the  paper  brought 
before  us  by  Dr.  Lewis  to  realize  the  real  solemnity  of  the  sub- 
ject that  we  are  handling. 


CLAUDE:   NEON  LIGHTING  371 

NEON  TUBE  LIGHTING.* 


BY  GEORGES  CLAUDE. 


It  is  known  that  the  discovery  of  the  rare  gases, — 
those  very  curious  gases  which  were  contained  in  the  atmosphere 
unknown  to  us — is  the  magnificent  work  of  Sir  W.  Ramsay.  *It 
is  known  also  that  it  was  the  distillation  of  liquified  air  which 
led  Ramsay  to  the  securing  of  such  wonderful  results,  results 
which  are  all  the  more  marvelous  since  they  were  obtained  with 
a  modest  apparatus  producing  from  1  to  2  liters  of  liquified  air 
per  hour. 

I  had  imagined  that  by  using  the  much  greater  facilities  at  my 
disposal  and  with  the  use  of  apparatus  which  can  liquify  10,000 
cubic  meters  of  air  per  day,  I  might  have  obtained  some  new 
results.  But  alas,  I  found  that  there  was  nothing  to  do  after 
Ramsay. 

But  if  in  spite  of  my  desires,  I  have  been  unable  to  add  to  the 
list  of  rare  gases,  I  have  nevertheless  been  able  to  produce  them, 
especially  so  in  the  case  of  the  neon,  in  far  larger  quantities. 
Things  are  so  arranged  in  my  oxygen  apparatus  that  this  neon  is 
the  residue  of  the  progressive  liquefaction  of  air,  and  is  ob- 
tained as  a  by-product  of  the  industrial  manufacturing  of  oxy- 
gen. The  output  of  this  apparatus  is  so  large,  that  in  spite 
of  the  insignificant  proportion  of  neon  in  the  air,  1  part  to 
66,000,  yet  with  a  modest  apparatus  of  50  cubic  meters  of  oxygen 
per  hour,  100  liters  of  neon  can  be  produced  in  a  day.  Balloons 
can  be  filled  with  this  gas  as  I  am  showing  it  to  you  here,  and 
balloons  that  can  fly,  as  the  density  of  neon  is  two-third  times 
that  of  air. 

Consequently,  neon  being  such  an  abundant  industrial  product, 
I  have  engaged  myself  in  a  search  for  uses  for  it. 

I  have  directed  my  efforts  toward  light  production.  I  do  not 
have  to  tell  you,  gentlemen,  what  a  serious  drawback  the  ever 
increasing,  dazzling  and  blinding  properties  of  modern  lighting 
possess.    And  you  are  all  aware  of  the  hopes  that  have  been  en- 

*  Outline  of  a  lecture  given  at  the  seventh  annual  convention  of  the  Illuminating 
Engineering  Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 


372  TRANSACTIONS  I.  E.   S. — PART  II 

tertained  for  the  uses  of  diffused  lighting  without  any  fatigue  to 
the  eyesight  and  which  might  be  secured  from  the  marvelous  and 
fine  brightness  with  which  the  rarefied  atmosphere  in  the  Geissler 
tubes  is  illuminated.  Unfortunately,  these  last  have  remained  up 
to  now  detestable  apparatuses,  since  the  luminescent  properties  of 
ordinary  gases  are  not  very  good.  Nitrogen,  is  the  only  gas  used, 
thanks  to  the  remarkable  perseverance  of  Moore,  although  the 
efficiency  of  his  apparatus  is  very  low — 1.7  to  2  watts  per  candle. 

Rare  gases  are  remarkable  for  their  ability  to  become  lumines- 
cent; their  spectra  are  remarkable;  the  one  of  neon  is  especially 
so.  It  contains  some  numerous  and  superb  lines  of  red,  orange 
and  yellow,  and  three  important  lines  of  green.  Unfortunately,  it 
contains  neither  white  nor  violet  tints.  Certainly,  this  absence  of 
blue  is  a  big  fault  so  far  as  a  source  of  lighting  is  concerned ;  but 
I  supposed  it  possible  to  correct  this  fault,  and  I  have  passed  on 
to  something  else. 

I  had  in  view  other  reasons,  besides  the  richness,  of  its  spec- 
trum, to  be  taken  up  in  a  study  of  neon.  First,  neon  really 
possesses  an  extraordinary  aptitude  to  become  luminous.  It  is  a 
long  time  since  the  illustrious  Sir.  J.  Dewar  has  succeeded 
in  the  production  of  tubes  with  neon  which  are  illuminated 
spontaneously  at  the  points  of  maximum  amplitudes  of  Hertz' 
interferences  and  which  are  extinguished  at  the  nodes.  So 
that  these  peculiar  detectors  furnish  very  original  means  with 
which  to  measure  the  length  of  the  waves  in  the  installations  of 
wireless  telegraphy. 

And  it  was  another  one  of  Ramsay's  colleagues,  Professor 
Collie,  who  with  the  neon  that  I  had  sent  to  Ramsay,  has  been 
able  to  make  the  following  observation,  a  very  important  one  in- 
deed: with  a  sealed  glass  tube  containing  a  little  quicksilver  in 
a  rarefied  atmosphere  of  neon,  if  one  shakes  the  tube  in  a  dark 
place,  the  mercury  looks  like  a  real  rain  of  fire.  This  curious 
phenomenon  is  explained  by  the  result  of  Bouty's  experiences. 
This  scientist  has  indeed  been  able  to  observe  with  great  surprise 
that  neon  is  easily  passed  through  by  an  electric  discharge :  where 
it  is  necessary  to  have  1,000  volts,  for  instance,  in  the  case  of  air, 
but  13  volts  are  sufficient  with  neon;  and  you  can  then  appre- 


CLAUDE:   NEON  LIGHTING  373 

ciate  Collie's  phenomenon:  the  electrifying  of  the  tube  by  the 
shaking  of  the  quicksilver  is  enough  to  cause  discharges  in  the 
gas. 

With  such  properties,  neon  can  very  well  enter  the  field  of 
production  of  luminescence.  I  have  therefore,  directed  my  ef- 
forts toward  using  this  gas  in  immense  Geissler  tubes,  similar  to 
the  Moore  tubes.  The  first  obstacle  encountered  was  indeed  a 
queer  one.  You  know,  gentlemen,  that  there  is  also  another  name 
for  rare  gases — noble  gases.  Now,  it  would  appear  that  neon  en- 
tertains a  very  lofty  idea  of  its  dignity;  it  is  quite  capable  of 
working  wonders  by  itself;  it  absolutely  refuses  to  perform  any- 
thing when  it  finds  itself  in  contact  with  those  inferior  fellows 
known  for  instance  as  hydrogen  or  nitrogen.  When  it  is  re- 
membered that  these  gases  are  much  more  inferior  than  it,  either 
in  their  aptitude  for  luminescence  or  with  respect  to  luminous 
efficiency,  even  minute  traces  of  them  mixed  with  neon  are  suffi- 
cient to  displease  it  to  such  an  extent  that  none  of  the  lines  of 
its  spectrum  are  visible  in  the  light  produced.  Here  is  a  tube 
which  has  99  per  cent,  of  neon  and  1  per  cent,  of  nitrogen.  Only 
the  light  produced  by  the  latter  appears. 

It  is  not  enough,  therefore,  to  introduce  into  the  luminescent 
tube  absolutely  pure  neon,  for  if  at  the  start  the  tube  shines  with 
a  magnificent  brightness,  the  impurities  emitted  by  the  electrodes 
when  the  current  is  passing  through,  cause  very  rapidly  the  drop 
of  its  luminescence.  In  order  to  overcome  this  serious  difficulty, 
I  was  compelled  to  devise  a  process  which  would  purify  the  neon 
in  the  tube  itself,  as  fast  as  the  impurities  were  introduced  into  it 
by  the  passage  of  the  current.  I  was  led  with  this  object  in  view, 
to  make  use  of  one  of  the  curious  resources  of  liquified  air;  the 
remarkable  property  discovered  by  Dewar  that  charcoal  absorbs 
air  with  great  energy  when  frozen  to  the  temperature  of  liquified 
air. 

But  it  is  under  peculiar  conditions  that  I  use  this  property. 
Charcoal  does  not  mingle  indifferently  with  every  and  all  gases. 
Generally  speaking  the  harder  they  are  to  liquify  the  harder  it  is 
to  absorb  them ;  this  absorption  is  very  much  smaller  in  the  case  of 
neon  than  with  the  different  sorts  of  gases  which  might  accom- 
pany it.    And  you  can  then  conceive  easily  the  process  which  I 


374  TRANSACTIONS  I.   E.   S. — PART  II 

have  devised  in  order  to  purify  the  atmosphere  of  my  neon  tubes. 
The  tube  to  be  rilled  is  connected  to  a  charcoal  receiver,  immersed 
in  liquified  air.  This  charcoal  slowly  absorbs  the  gases  developed 
through  the  passage  of  the  current,  but  it  leaves  the  neon.  By 
this  contrivance  a  pump  has  been  made;  but  this  is  an  intelligent 
pump  which  sucks  and  carries  away  the  troublesome  molecules, 
and  respects  the  others.  After  a  laborious  process  which  lasts 
for  many  hours,  the  neon  remains  victorious ;  the  tube  is  finished, 
sealed  off  from  the  charcoal  receiver,  and  will  show  thereafter 
without  weakening  the  superb  light  of  the  neon.  Superb !  well 
all  tastes  are  different  and  you  will  perhaps  find  that  I  exagger- 
ate; but  I  shall  try  just  now  to  give  you  a  more  complete  satis- 
faction. 

This  difficulty  having  been  overcome,  another  one  cropped  up. 
I  had  noticed  that  the  neon  tubes  thus  obtained  were  short-lived. 
After  showing  a  rapid  increase  in  the  difference  of  potential 
at  the  bars,  they  began  to  flicker  and  to  crackle,  and  lastly  went 
out  in  5  or  6  hours.  Well,  gentlemen,  you  will  easily  recognize 
there  the  strange  phenomenon  discovered  by  Moore.  Moore  has 
observed  indeed  that  the  atmosphere  inside  of  his  tubes  was  rare- 
fying itself  progressively  and  that  finally  the  tubes  went  out  en- 
tirely; this  fact,  a  very  mysterious  one  indeed,  had  neutralized 
his  efforts  up  to  the  moment  when  he  had  the  idea  of  introducing 
nitrogen  in  his  tubes,  by  means  of  an  ingenious  electromagnetic 
valve,  as  fast  as  the  rarefaction  took  place. 

Unfortunately,  such  a  remedy  could  not  be  applied  in  the  case 
of  neon  for  if  Moore's  observations  are  accurate,  the  quantity 
of  nitrogen  absorbed  by  his  tubes  is  astonishing:  200  liters  per 
year  for  a  tube  of  50  m.  If  my  neon  tubes  were  such  gluttons, 
my  apparatus  of  liquified  air  would  be  inadequate  to  meet  their 
demand.  It  was  of  course  necessary  for  me  to  look  for  condi- 
tions which  would  permit  me  not  to  consider  the  absorption  of 
neon,  so  that  one  charge,  one  single  dose  of  neon,  could  insure 
to  the  tube  a  very  long  life,  one  comparable  to  that  of  an  incan- 
descent lamp. 

To  attain  this  result,  it  was  necessary  to  begin  by  determining 
the  manner  of  this  absorption.  I  was  able  to  observe  at  the 
outset  that  the  electrodes  of  the  first  tubes,  which  were  very 


CLAUDE:   NEON  LIGHTING  375 

small,  became  incandescent  upon  the  passage  of  the  current  and 
volatilized  rapidly.  A  metallic  deposit  used  to  form  from  these 
scales  and  strips  in  the  neighborhood  of  the  electrodes.  I 
have  thought  that  it  was  this  metallic  deposit  which  on  forming 
absorbed  the  neon.  And  in  fact,  by  dissolving  these  deposits  in 
nitric  acid,  gases  were  developed  containing  neon. 

Therefore,  there  is  no  doubt  that  the  volatilization  of  the 
electrodes  is  what  makes  the  trouble.  In  order  to  minimize  it,  the 
use  of  large  electrodes  will  be  necessary  henceforth,  which  can 
be  very  little  volatilized  by  the  current.  Experience  has  confirmed 
this  supposition  and  to  such  an  extent  that  by  using  electrodes  of 
5  sq.  dm.  per  ampere  the  volatilization  is  rendered  nil  and  the  life 
of  the  tubes  lengthened  considerably. 

The  life  increases  naturally  with  the  length  of  the  tubes,  as 
there  are  always  only  two  electrodes  to  absorb  the  neon,  and  that 
the  longer  the  tube  the  more  neon  it  contains.  With  tubes  6  m. 
long,  you  can  attain  easily  1,000  or  1,200  hours,  and  I  have  tubes 
of  20  m.  in  excellent  condition  after  2,000  hours.  This  is  super- 
ior to  the  incandescent  lamps. 

Here  is  the  problem  solved  then,  and  solved  in  a  manner  unex- 
pectedly simple.  We  are  in  possession  of  tubes  capable  of  show- 
ing the  spectrum  of  neon  in  all  its  purity.  These  tubes,  which  are 
not  provided  with  valves,  are  much  more  simple  than  those  prev- 
iously made,  and  give  an  entirely  satisfactory  length  of  life. 

What  advantages  have  these  tubes  as  compared  with  the  nitro- 
gen tubes?  Gentlemen,  I  shall  only  insist  upon  the  essential 
points  and  any  one  desiring  further  details  is  referred  to  the 
paper  which  I  read  before  the  Societe  des  Electriciens,  November 
8,  1911. 

At  the  outset  the  necessary  difference  of  potentials  is  three 
times  less  than  with  nitrogen ;  this  is  a  great  advantage  with  re- 
spect to  safety.  These  6  m.  tubes  have  less  than  800  volts  at 
their  ends.  Three  could  be  mounted  upon  a  transformer  of  3,500 
volts. 

Secondly,  its  illuminating  power  is  higher,  200  candles  per 
meter  instead  of  60.  You  can  consequently  use  for  lighting  tubes 
which  are  much  shorter  and  accordingly  less  expensive.    Another 


3/6  TRANSACTIONS  I.   E-   S. — PART  II 

advantage  to  be  derived  is  that  these  tubes  can  be  manufactured 
in  factories  and  carried  all  ready  for  use  to  a  customer  or  client. 

And  lastly — this  is  the  most  important  part — the  illuminating 
efficiency  is  much  greater.  Instead  of  1.7  watt  per  candle  ob- 
tained by  the  use  of  nitrogen,  only  0.5  is  required  for  the  long 
neon  tubes.  May  be  you  do  not  consider  this  marvelous  when  you 
make  a  comparison  with  arc  lights ;  but  it  has  to  be  borne  in 
mind  that  the  question  is  one  of  spheric  and  not  hemispheric 
watts.  Furthermore,  no  expensive  carbons  are  required  and  all 
upkeep  charges  are  done  away  with.  Really,  when  everything  is 
considered,  I  believe  that  with  the  exception  of  the  mercury 
lamps,  the  neon  tubes  supply  the  most  economical  lighting. 

In  some  interesting  experiments  carried  out  at  the  Laboratoire 
Central  d'Electricite,  Messrs.  Broca  and  Laporte  have  observed 
that  the  neon  light  is  physiologically  excellent  on  account  of  its 
dull  luminescence  and  that  it  increases  visual  acuity  by  25 
per  cent.  You  can  really  notice  with  what  clearness  and  sharp- 
ness the  small  figures  on  the  reports  which  I  am  passing  around 
can  be  seen. 

All  is  rather  perfect  therefore  but  for  the  color,  which  is  an- 
other matter.  Evidently  this  light  is  too  red ;  it  is  too  red  because 
of  its  want  of  blue.  Look  at  this  bouquet ;  it  is  of  a  beautiful  blue ; 
see  how  dull  and  disappointing  it  is.  On  the  contrary  here  are 
some  poppies :  see  how  resplendent  they  appear.  No  doubt,  this 
predominance  of  the  red  color  allows  of  some  beautiful  illuminat- 
ing effects.  Here  you  have  as  an  instance,  the  Grand  Palais  at 
the  Champs  Elysees,  in  Paris,  lighted  by  neon  in  1910  on  the 
occasion  of  the  automobile  show  held  then;  and  the  St.  Ouen 
Church  of  Rouen  which  was  lighted  by  50  neon  tubes  during  the 
festival  of  the  Norman  millenium.  Undoubtedly,  in  a  number 
of  cases  of  industrial  lighting  this  light  could  be  applied  in 
preference  to  that  of  mercury  light,  since  it  is  very  economical. 
Allow  me  indeed  to  insist  and  to  bring  to  your  notice  with  what 
strange  facility  you  have  accustomed  yourselves  to  such  a  red 
light,  so  as  to  retain  only  a  pleasantly  warm  impression  of  a 
golden  yellow  from  which  the  red  is  completely  absent. 

But  I  have  to  admit  that  this  excess  of  red  is  hardly  acceptable 
in  the  majority  of  cases  and  I  have  applied  myself  to  having 


CLAUDE:    NEON   LIGHTING  377 

this  light  corrected.  There  is  a  solution  which  seems  to  me  plain 
and  that  is  to  combine  the  pale  mercury  light  with  that  of  the 
bright  neon.  In  this  case,  however,  there  are  two  difficulties  to 
be  overcome.  The  first  is  that  mercury  and  neon  when  placed  in 
the  same  tube  will  not  blend  and  work  together;  second  is  that, 
if  it  be  required  to  mingle  blue  tubes  with  red  tubes,  the  Cooper- 
Hewitt  mercury-vapor  tubes  would  require  a  continuous  current 
at  a  low  tension,  whereas  the  neon  tubes  demand  alternating  cur- 
rent at  high  tension.  These  do  not  go  together.  I  have,  however, 
made  correcting  tubes,  which  are  similar  to  the  ordinary  neon 
tubes  but  having  a  little  mercury.  These  tubes  light  with  the 
alternating  current  as  in  the  case  of  usual  neon  tubes;  but  mer- 
cury volatilizes  progressively  and  the  blue  light  which  mercury 
gives,  invades  the  whole  tube. 

Well,  here  is  progress,  to  be  sure.  Our  blue  bouquet  has 
recovered  its  colors,  but  it  is  our  poppy  bouquet  which  now  looks 
pitiable ;  and  as  to  ourselves,  gentlemen,  instead  of  being  rubi- 
cund, we  are  now  just  ghastly  pale. 

Patience!  I  light  this  neon  tube  and  here  we  have  the  sun's 
light  succeeding  to  that  of  the  moon's  pale  light.  See  how  every- 
thing has  returned  to  its  normal  state;  the  blue  color  of  this  bou- 
quet, the  red  of  that  one,  the  delicate  hues  of  those  flowers  so 
varied ;  and  above  all  you  have  noticed,  ladies,  how  your  complex- 
ion matches  nicely  with  this  light.  And  what  interesting  orna- 
mental effects  can  be  obtained  by  the  combination  of  tubes  of 
different  colors.  Here  is  an  example,  a  fixture  put  up  by  the 
firm  of  Paz  &  Silva  for  the  automobile  show  in  Paris. 

The  efficiency  of  these  correcting  mercury  tubes  is,  unfortu- 
nately, notably  inferior  to  that  of  the  neon  tubes  and  should  be 
in  the  near  neighborhood  of  i  watt  per  candle.  It  is  acceptable 
even  at  that.  Having  an  equal  number  of  blue  and  red  tubes, 
you  will  see  that  a  very  pleasant  light  is  obtained,  very  much  dif- 
fused, without  shadows  and  at  an  energy  consumption  of  0.8 
watt  per  candle.  This  is  better  than  what  has  hitherto  been 
obtained   with   luminescent  lighting. 

Gentlemen,  I  have  another  application  to  bring  to  your  notice. 
If  the  objectionable  red  features  in  this  light  cannot  be  always 
neutralized,  there  are  cases  where  it  proves  to  be  an  unquestion- 


37&  TRANSACTIONS  I.  E.   S. — PART  II 

able  advantage.  To  begin  with,  for  illumination  of  monuments, 
as  I  have  already  remarked;  but  it  is  of  inestimable  value  for 
advertising  illumination,  where  the  more  dazzling  the  light  is, 
the  more  it  will  strike  the  eye,  and  hence  the  better  it  will  be. 
Now,  with  neon  you  are  liberally  served. 

I  have  been  able  to  make  with  the  aid  of  my  collaborator,  M. 
de  Beaufort,  some  tubes  of  a  small  diameter  which  can  be  bent, 
or  twisted  without  difficulty  and  be  given  any  desired  form,  and 
be  lighted  with  red  or  blue  lights.  There  might  be  some  appre- 
hension that  the  minute  quantity  of  neon  contained  in  those  tubes 
might  give  them  only  a  short  life.  However,  I  had  the  pleasant 
surprise  of  finding  out  that  with  the  sole  condition  respecting 
the  rule  of  the  feeble  density  of  the  current  at  the  electrodes, 
these  small  tubes  lasted  as  long  as  the  big  ones.  Here  is  a  tube 
which  has  burned  1,400  hours.  This  one  operates  at  the  rate  of 
30  milli-amperes,  sufficient  as  you  can  see,  to  give  to  it  a  very 
luminous  aspect. 

These  small  apparatuses  work  on  a  common  transformer  with 
alternating  current.  Upon  continuous  current,  the  transformer 
is  controlled  by  a  rotating  or  Wehnelt  interrupter.  The  cost  of 
an  installation  is  not  much  higher  than  that  of  ordinary  apparatus 
and  the  consumption  of  energy  is  less  and  affords  a  much  better 
effect.  There  is  certainly  to  be  found  here  a  brand  new  method 
for  the  industrv  of  luminous  advertising. 


ROSE  :   LABORATORY   OF   GENERAL   ELECTRIC    CO.  3/9 

THE  ILLUMINATING  ENGINEERING  LABORATORY 
OF  THE  GENERAL  ELECTRIC  COMPANY.* 


BY  S.  L.   E.  ROSE. 


Synopsis:  About  the  year  1895  the  General  Electric  Company  started 
at  the  Lynn  works  the  study  of  illumination  problems  and  the  proper 
application  of  arc  lamps.  In  1909,  the  department  was  moved  from  Lynn 
to  Schenectady.  The  present  laboratory  is  equipped  for  making  tests  on 
all  kinds  of  illuminants.  The  work  of  the  laboratory  is  divided  into 
four  main  divisions:  namely;  commercial  investigations  and  their  appli- 
cations, photometric  testing  and  developmental,  research,  and  photo- 
graphic. Facilities  are  available  for  a  thorough  investigation  of  all  means 
and  methods  of  artificial  illumination  and  the  testing  of  lighting  units  for 
commercial  or  special  work.  Parts  of  the  laboratory  equipment  are 
described  and  illustrated  in  this  paper. 

Before  giving  a  description  of  this  illuminating  engineering 
laboratory  and  the  work  being  carried  on  there,  perhaps  a  review 
of  its  growth  from  its  inception  to  its  present  proportions  will 
be  of  interest. 

About  the  year  1895,  the  General  Electric  Company  started 
at  its  Lynn  Works  the  study  of  illumination  problems  and  the 
proper  application  of  arc  lamps.  This  was  the  beginning  of  what 
is  now  known  as  the  illuminating  engineering  laboratory  and  the 
work  was  carried  on  under  the  direction  of  Mr.  W.  D'A.  Ryan 
who  is  director  of  the  present  laboratory.  The  first  photometer 
room  occupied  a  floor  space  of  about  500  square  feet  (38.09 
sq.  m.).  The  photometer  was  of  the  single-mirror  crane  type 
and  could  be  used  as  a  constant  length,  constant  intensity  or 
constant  radius  photometer.  All  photometric  testing  was  done 
on  this  one  photometer  and  it  was  a  number  of  years  before  more 
space  was  devoted  to  this  work. 

In  the  fall  of  1907,  three  rooms,  wi.th  an  aggregate  floor  space 
of  approximately  1,500  square  feet  (114.27  sq.  m.),  were  built 
in  a  new  factory  building  and  three  photometers  installed,  one  for 
small  units,  one  for  large  units  and  one  for  miscellaneous  work. 

*A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 


38o 


TRANSACTIONS  I.  E.   S. — PART  II 


In  1909  the  illuminating  engineering  department  was  moved 
from  Lynn  to  Schenectady.  None  of  the  photometers  was 
moved  from  Lynn  and  the  ones  now  in  use  at  Schenectady  were 
designed,  built  and  installed  under  the  supervision  of  the  depart- 
ment. A  description  of  the  photometers  used  at  Lynn  and  the 
ones  in  use  at  Schenectady  for  large  unit  work  as  well  as  the 
methods  of  test  have  been  given  in  a  previous  paper  before  this 
society.1 

The  present  laboratory  is  situated  in  a  two-story  brick  building 
with  an  aggregate  floor  space  of  approximately  7,000  square  feet 
(533.27  sq.  m.).     (See  Fig.  1.) 


DARK  /    PHOTOGRAFIC 
ROOM    I         DEPT. 


SHOW 
LECTURE 

AND 
DRAFTING    M 

ROOM 


SHOW-ROOM 


a 

ULBRICHT 
SPHERE 
2M.0IA. 


SECOND  FLOOR  PLAN 


GENERAL 
OFFICE 


5  /< 


RESEARCH  AND 
CALIBRATION  ROOM 


ULBRICHT 
SPHERE 
IM.DIA 


AT 


O' 


\     <£■: 


WORK  ROOM 


FIRST  FLOOR   PLAN 
Fig.  1. — Plan  of  laboratory. 


:oal  ) 


Two  universal  rotators  and  a  single-mirror  crane  type  pho- 
tometer (Fig.  2)  are  available  for  obtaining  candle-power  distri- 
bution curves  of  small  units.  The  latter  photometer  is  very  easily 
changed  into  a  constant  length,  constant  intensity  or  constant 
radius  photometer.  It  is  mostly  used  as  a  constant  radius  pho- 
tometer and  the  radius  of  test-may  be  changed  at  will  from  5  feet 
(1.524  m.)  to  20  feet  (6.096  m.)  or  any  intermediate  point  accord- 
ing to  the  size  of  the  unit  being  tested.    For  convenience  of  cal- 

1 1.  E.  S.  Trans.,  page  641,  vol.  6,  1911. 


ROSE:    LABORATORY   OF   GENERAL   ELECTRIC   CO. 


38l 


culation,  all  tests,  so  far  as  possible,  are  made  at  a  radius  of  10 
feet  (3.04  m.).  A  sector  disk  is  employed  to  increase  the  range 
of  the  photometer,  when  necessary.  For  obtaining  total  flux 
on  small  units,  absorption  of  small  units,  globes  or  balls,  an 
Ulbright  sphere  of  1  meter  diameter  is  used  and  for  large  units 
an  Ulbright  sphere  of  2  meters  diameter  is  employed.  In  addi- 
tion to  the  regular  photometers  mentioned  above,  a  number  of 
portable  photometers,  of  both  foreign  and  domestic  make,  are 
available  for  outside  tests  and  for  special  work  in  the  laboratory. 
For  spectrum  analysis,  color  absorption,  research  and  special 
investigations,  a  spectrophotometer,  spectrometer  and  colorimeter 
are  available. 


Fig.  2. --Single-mirror  crane  type  photometer  for  small  light  units. 


Energy  for  lighting,  power  and  experimental  purposes  is  avail- 
able from  the  direct  current  three-wire  shop  circuit  of  the  General 
Electric  Company,  the  alternating  current  three-wire  commercial 
city  circuit,  a  motor  driven  alternator  which  can  be  arranged  to 
give  25  to  60-cycle  current,  two  motor  driven  direct  current  gene- 
rators of  500  to  700  volts  and  125  volts  respectively,  and  a 
constant  current  transformer  for  alternating  current  series  work. 
A  switchboard  panel  is  so  arranged  that  current  from  any  of  the 
above  sources  may  be  switched  on  any  circuit  in  the  building.  In 
addition  to  this,  a  circuit  leads  to  each  room  from  a  60-cell  storage 


382  TRANSACTIONS  I.   E.   S. — PART  II 

battery.  This  battery  is  primarily  for  furnishing  energy  for 
incandescent  unit  work  and  working  standard  lamps,  but  may  be 
used  as  an  emergency  supply,  if  necessary. 

For  the  testing  of  gas  units,  the  laboratory  is  equipped  for  both 
high  and  low  pressure  work.  For  measuring  consumption,  wet 
test  meters  are  available  which  will  operate  on  pressures  from 
2  to  3  inches  (5.08  cm.  to  7.62  cm.)  of  water  up  to  10  pounds 
(4.53  kilograms)  per  square  inch  (6.45  sq.  cm.).  A  motor  driven 
compressor  is  installed  which  will  take  gas  from  the  city  mains 
and  deliver  it  at  pressures  ranging  from  2  or  3  inches  (5.08  or 
7.62  cm.)  of  water  up  to  10  pounds  (4.53  kilograms)  per  square 
inch  (6.45  sq.  cm.).  Indicating  and  recording  pressure  gauges, 
an  indicating  gravitometer  and  a  recording  calorimeter  complete 
the  equipment.     (Fig.  3.) 

The  work  of  the  laboratory  is  divided  into  four  main  divisions, 
namely :  commercial  investigations  and  applications,  photometric 
testing  and  developmental,  research  and  photographic.  The 
whole  comprises  a  thorough  investigation  of  all  means  and  meth- 
ods of  artificial  illumination  and  testing  of  lighting  units  for 
commercial  or  special  work. 

The  commercial  division  is  constantly  giving  advice  and  fur- 
nishing lighting  recommendations  to  all  parts  of  the  world  and 
for  all  classes  of  lighting  of  which  the  following  may  be  men- 
tioned as  an  illustration  of  the  diversified  character  of  the  work ; 
the  office  building  of  the  Buffalo  General  Electric  Company,2  the 
Panama  Canal,  and  the  Panama-Pacific  International  Exposition 
to  be  held  in  San  Francisco  in  1915.  In  connection  with  the 
commercial  division,  a  display  of  various  types  of  lighting 
units  is  maintained  together  with  intensity  and  color  booths 
(Fig.  4).  The  latter  two  are  probably  the  demonstrations  of 
greatest  interest  to  the  public.  The  terms  used  by  the  expert  in 
the  art  of  illuminating  engineering  often  seems,  to  the  layman, 
vague  and  inexpressive.  The  merchant  may  be  told,  by  the  engi- 
neer, that  he  requires  for  proper  illumination  of  different  depart- 
ments of  his  store,  2,  5,  7  or  10  foot-candles,  but  owing  to  the 
vagueness  of  his  conception  of  the  foot-candle,  he  is  still  neces- 
sarily "in  the  dark,"  as  to  the  amount  of  illumination  he  is  con- 
tracting to  buy. 

2  For  full  description  see  I.  E-  S.  Trans.,  vol.  VII,  page  597,  1912. 


Fig.  3.—  Gravitonieter  and  calorgraph. 


3* 


t 


Fig.  4.  — Demonstration  room. 


rose:  laboratory  of  general  electric  co.  383 

The  idea  of  the  intensity  booths  is  to  show  the  intensity  of 
illumination  in  steps  of  }4,  Yz,  l\  3.  5»  7^>  IO>  :5  an(i  20  foot- 
candles.  This  demonstration  is  made  in  a  row  of  booths  extend- 
ing along  one  side  of  a  room,  each  booth  measuring  approximately 
2  ft.  by  3  ft.  6  in.  (0.61  X  1.07  m.).  By  the  manipulation  of 
switches,  these  same  booths  are  used  to  demonstrate  the  differ- 
ence in  color  of  the  ordinary  illuminants  now  in  use  and  daylight. 

The  photometric  division  furnishes  illumination  data  to  all  the 
other  divisions,  outside  departments,  sales  offices  and  through 
them  to  the  general  public.  Thorough  tests  are  made  on  all  kinds 
of  lamps  and  lighting  equipment.  Street  lighting  and  interior 
systems  are  tested  under  operating  conditions.  Experimental 
tests  are  carried  on  night  and  day  when  necessary  to  try  out 
some  new  piece  of  apparatus  for  the  patent  department,  or  fur- 
nish special  data  to  the  engineers  of  the  laboratory.  New  designs 
of  lighting  apparatus  and  systems  are  constantly  being  devised, 
constructed  and  tried  out  to  determine  their  commercial  value  or 
their  application  to  some  special  purpose.     (See  Fig.  6.) 

The  research  division  carries  on  special  investigations  of  a 
scientific  nature. 

The  photographic  division  furnishes  the  commercial  division 
as  well  as  the  sales  offices,  architects,  engineers,  etc.,  with  day 
and  night  views  of  representative  installations  as  well  as  many 
conditions  illustrating  problems  which  are  encountered  and 
examples  of  both  good  and  bad  lighting. 

The  purpose  of  the  laboratory  is  utilitarian  and  altruistic  and 
it  is  devoted  to  the  services  of  producers  and  consumers  of  arti- 
ficial light  and  to  the  betterment  of  the  art  of  illumination. 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 

Published  monthly,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIII 


NOVEMBER.  1913 


No.  8 


Council  Notes. 

A  meeting  of  the  Council  was  held  in 
the  general  offices  of  the  society,  29 
West  39th  Street,  New  York,  November 
14,  1913.  Those  present  were:  C.  O. 
Bond,  president;  Joseph  D.  Israel,  gen- 
eral secretary;  V.  R.  Lansingh,  C.  A. 
Littlefield,  L.  B.  Marks,  treasurer; 
Preston  S.  Millar,  C.  J.  Russell,  F.  J. 
Rutledge,  W.  J.  Serrill  and  G.  H. 
Stickney. 

The  meeting  was  called  to  order  at 
10:30  a.  M.  by  President  C.  O.  Bond. 

The  minutes  of  the  October  meeting 
were  adopted. 

Mr.  C.  A.  Littlefield,  chairman  of  the 
Finance  Committee,  reported  that  his 
committii  had  approved  vouchers  Nos. 
1476  to  15 12  inclusive,  aggregating 
$1,646.25.  Payment  of  these  vouchers 
was  authorized.  The  committee  also 
recommended  an  appropriation  of  $100 
as  the  quota  of  the  society  towards  the 
expenses  of  an  International  Commis- 
sion on  Illumination.  The  appropriation 
was  granted.  Mr.  Littlefield,  also  re- 
ported that  the  Finance  Committee 
would  submit  at  the  December  Council 
meeting  a  budget  for  the  present  admin- 
istration. 

A  report  on  the  accounts  of  the  society 
for  the  year  of  January  1  to  September 
30,  1913,  which  had  been  prepared  by  a 
certified  public  accountant,  at  the  re- 
quest of  the  Finance  Committee  of  the 


previous  administration,  was  received. 
The  report  will  be  published  in  the 
Transactions.  The  earnings  for  that 
period  amounted  to  $8,256.49;  while  the 
expenses,  including  an  estimate  of  out- 
standing debts,  aggregated  $8,175.65. 

Mr.  Israel  reported  that  the  total 
membership  of  the  society,  counting  the 
resignations  and  applications  presented 
at  the  meeting  was  1,392,  and  that  the 
expenditures  for  the  first  month  of  the 
present  fiscal  year  had  aggregated 
$2,893.17.  The  receipts  during  that 
period  amounted  to  $291.22. 

The  following  amended  report  of  the 
Council  Executive  Committee,  covering 
business  transacted  by  the  committee 
since  the  previous  Council  meeting,  was 
adopted : 

A  meeting  of  the  Council  Executive  Com- 
mittee was  held  in  the  general  offices  of  the 
society,  October  31,  1913.  Those  present  were 
Chas.  O.  Bond,  president;  Joseph  D.  Israel, 
general  secretary;  C.  A.  Littlefield,  L.  B. 
Marks,   treasurer,    and   Preston    S.    Millar. 

Mr.  Preston  S.  Millar  was  appointed  chair- 
man of  a  Committee  on  Education  of  the 
Illuminating  Engineering  Society.  The  former 
name  of  this  committee  was  Committee  on 
Collegiate  Education.  It  was  suggested  that 
among  other  things  the  committee  be  asked 
to  consider  the  preparation  of  a  tentative 
course  in  illuminating  engineering  for  schools 
and  colleges. 

Mr.  C.  E.  Clewell  was  appointed  chairman 
of  a  sub-committee  on  Office  Lighting  of  the 
Committee  on  Popular  Lectures.  An  appro- 
priation of  $25  was  granted  for  stenographic 
expenses  in  connection  with  the  work  of  the 
sub-committee. 


TRANSACTIONS    I.    E.    S.—  PART    I 


The  following  additional  committee  appoint- 
ments were  made:  Finance:  A.  Hertz,  W.  J. 
Serrill;  Section  Development:  Joseph  Langan; 
Membership:  C.  J.  Ramsburg;  Board  of  Ex- 
aminers: W.  Cullen  Morris,  chairman,  and 
C.  H.  Sharp;  Editing  and  Publication:  Clay- 
ton H.  Sharp,  chairman;  A.  S.  McAllister, 
VV.   J.    Serrill. 

Twelve  applicants  were  elected  mem- 
bers. Their  names  appear  on  another 
page. 

Ten  resignations  were  accepted. 

The  following  report  pertaining  to 
the  society's  policy  in  supplying  copies 
of  papers  to  authors  was  received  and 
adopted : 

Your  Committee,  appointed  to  reconsider  the 
policy  of  the  society  in  supplying  authors 
with  copies  of  their  papers  which  are  pre- 
sented before  meetings  of  the  society,  begs 
leave  to  submit  the  following  recommendations: 

When  a  paper  is  printed  in  advance,  twenty- 
five   (25)   copies  shall  be  sent  to  the  author. 

When  an  author  furnishes  to  the  general 
office  the  names  and  addresses  of  ten  (10)  or 
less,  non-members  of  the  society,  to  whom  he 
desires  to  have  copies  of  his  paper  sent,  the 
general  office  will,  if  copies  are  available,  mail 
to  each  of  these  persons  a  copy  of  the  Trans- 
actions containing  the  paper  in  question,  indi- 
cating at  whose   request  it  is  sent. 

That  the  standard  form  of  reprint  be 
changed  to  secure  minimum  cost  compatible 
with  good  appearance;  the  pagination  of  the 
paper,  as  printed  in  the  Transactions,  being 
considered  satisfactory.  This  will  not  pre- 
clude the  issuing  of  more  expensive  reprints 
when  so  ordered. 

That  a  suitable  notice  disclaiming  responsi- 
bility of  the  society  for  statements  or  opinions 
of  authors  be  printed  on  the  title  page  of  all 
papers,  whether  in  the  Transactions,  advance 
copies   or   reprints. 

Respectfully  submitted, 
Herbert  E.   Ives, 
C.  H.  Sharp, 
G.    H.    Stickney,    Chairman. 

The  report  was  adopted  and  a  vote  of 
thanks  extended  to  the  committee. 

In  accordance  with  a  recommendation 
in  the  foregoing  report  it  was  voted  to 
place  the  following  statement  on  all 
papers  of  the  society : 


The  Illuminating  Engineering  Society  is  not 
responsible  for  the  statements  and  opinions 
advanced   by   contributors. 

Mr.  V.  R.  Lansingh  reported  verbally 
on  the  progress  of  the  work  of  his  Com- 
mittee on  Sustaining  Afembership. 

Reports  on  section  activities  during 
the  past  month  were  received  from  the 
following  vice-presidents :  Mr.  G.  H. 
Stickney  representing  New  York,  Mr. 
W.  J.  Serrill  representing  Philadelphia 
and  Mr.  J.  W.  Cowles  representing  New 
England. 

Mr.  Israel  reported  on  the  activities  of 
the  Pittsburgh  and  Chicago  Sections. 

The  following  appointments  to  com- 
mittees were  confirmed :  Nomenclature 
and  Standards,  A.  E.  Kennelly,  chair- 
man; Research,  H.  E.  Ives,  chairman; 
Advertising,  R.  E.  Campbell;  Papers, 
E.  J.  Edwards,  George  S.  Barrows, 
H.  A.  Hornor,  C.  E.  Stephens,  M.  G. 
Lloyd,  Alexander  Duane;  Reciprocal 
Relations,  W.  J.  Serrill,  chairman; 
Progress,  T.  J.  Litle,  Jr.,  E.  L.  Elliott, 
T.  W.  Roth,  W.  E.  Wickenden,  H.  S. 
Hower,  Wendell  Reber ;  Glare  from 
Reflecting  Surfaces,  F.  A.  Vaughn, 
N.  M.  Black,  J.  R.  Cravath,  F.  H.  Gil- 
pin, M.  G.  Lloyd;  Lighting  Legislation, 
Ellice  M.  Alger,  Oscar  H.  Fogg,  Her- 
bert E.  Ives,  Clarence  L.  Law,  F.  J. 
Miller,  G.  H.  Stickney,  L.  A.  Tanzer, 
W.  H.  Tolman. 

It  was  resolved  that  the  Council  bf 
the  Illuminating  Engineering  Society 
extend  a  vote  of  thanks  to  Mr.  Joseph 
B.  Gregg  for  his  valuable  assistance  in 
arbitrating  an  account  presented  by  the 
Hill  Publishing  Company. 

Informal  mention  was  made  of  a  re- 
cent movement  started  in  the  West  to 
organize  a  San  Francisco  or  a  Pacific 
Coast  Section  of  the  society. 

The   meeting  was   adjourned   at    1  :i5 

P.M. 


TRANSACTIONS   I.    E.    S.— PART   I 


Section  Notes. 


CHICAGO    SECTION 


A  meeting  of  the  Board  of  Managers 
was  held  in  the  Grand  Pacific  Hotel, 
November  5.  Those  present  were :  Dr. 
M.  G.  Lloyd,  chairman;  J.  B.  Jackson, 
secretary;  J.  R.  Cravath,  M.  J.  Sturm, 
and  H.  B.  Wheeler.  After  considerable 
discussion  of  a  program  for  the  present 
year,  it  was  decided  that  it  was  not 
desirable  to  arrange  a  definite  program 
for  the  full  year,  on  account  of  the 
arrangement  of  a  number  of  joint  meet- 
ings with  other  societies,  the  dates  of 
which  could  not  be  determined  at  the 
present.  Mr.  Sturm  was  delegated  to 
arrange  a  joint  meeting  with  the 
Chicago  Architectural  Business  Men's 
Association  for  January. 

A  meeting  of  the  Chicago  Section  was 
held  in  the  Auditorium  of  the  Western 
Society  of  Engineers,  Monadnock  Block. 
Chicago,  November  12.  A  paper  entitled 
"The  Illumination  of  Street  Railway 
Cars"  was  presented  by  Messrs.  L.  C. 
Porter  and  V.  L.  Staley  of  the  General 
Electric  Co.  An  excellent  exhibit  of 
reflectors,  holders,  fittings  and  accesso- 
ries was  arranged  by  the  Exhibition 
Committee,  of  which  Mr.  H.  B.  Wheeler 
is  chairman.  Mr.  W.  A.  Durgin,  Assist- 
ant Chief  Testing  Engineer  of  the  Com- 
monwealth Edison  Company,  gave  the 
first  of  a  series  of  20  minute  talks  on 
the  "Fundamentals  of  Illumination." 
Sixty-four  members  and  guests  attended 
the  meeting. 

The  December  meeting  is  to  be  held 
at  the  residence  of  Mr.  W.  A.  D.  Curtis, 
when  a  paper  entitled  "The  Lighting  of 
the  Home"  will  be  discussed. 

In  January,  there  will  be  a  joint  meet- 
ing with  the  Chicago  Architectural  Busi- 
ness Men's  Association,  at  which  Mr. 
J.  B.  Jackson,  secretary  of  the  Chicago 


Section,     will     give     a     paper     entitled 
"Planning  Lighting  Installations." 

NEW    ENGLAND    SECTION 

A  meeting  of  the  Board  of  Managers 
of  the  New  England  Section  was  held 
in  the  Hotel  Georgian,  November  3. 
Those  present  were :  C.  A.  B.  Halvor- 
son,  chairman;  C.  M.  Cole,  secretary; 
J.  \Y.  Cowles,  vice-president;  R.  B. 
Hussey,  J.  M.  Riley  and  R.  C.  Ware. 

The  meeting  was  devoted  to  the  con- 
sideration of  a  program  of  meetings  for 
the  present  year.  A  campaign  is  to  be 
undertaken  to  increase  the  membership 
of  the  section. 

NEW    YORK    SECTION 

A  meeting  of  the  Board  of  Managers 
was  held  November  15,  in  the  general 
office  of  the  society,  29  West  39th  Street, 
New  York  City.  Those  present  were: 
W.  C.  Morris,  chairman;  G.  H.  Stick- 
ney,  vice-president;  O.  H.  Fogg,  W.  H. 
Spencer,  H.  B.  McLean,  M.  D. 
McDonald,  H.  B.  Rogers,  S.  W.  Van 
Rensselaer,  and  C.  L.  Law,  secretary. 
The  meetings'  program  for  the  rest  of 
the  year  was  discussed.  Arrangements 
have  been  made  for  a  number  of  excel- 
lent papers  and  joint  meetings  with 
other  societies. 

The  New  York  Section  held  a  joint 
meeting  with  the  New  York  Companies' 
Section  of  the  National  Electric  Light 
Association  in  the  Auditorium  of  the 
New  York  Edison  Company,  November 
17.  Mr.  S.  G.  Rhodes  of  the  New  York 
Edison  Company  gave  a  talk  on  "Street 
Lighting  Abroad."  Mr.  H.  W.  Jackson 
of  the  General  Electric  Company  gave 
a  lecture  on  the  latest  improvements  in 
incandescent  lamps.  Mr.  Alexander 
Maxwell  of  the  New  York  Edison  Com- 
pany exhibited  several  Neon  tube  lamps. 
About  125  members  of  both  organiza- 
tions were  present. 


TRANSACTIONS    I.    E.    S. — PART    I 


PHILADELPHIA    SECTION 

On  November  20,  a  joint  meeting  of 
the  Philadelphia  Section  was  held  with 
an  ophthalmological  society  of  Philadel- 
phia. 

The  following  is  the  program  for 
future  meetings : 

Monday,  December  8. 
Joint  Meeting  with  Philadelphia  Section 

A.  I.  E.  E. 
"Brightness   Measurements   versus   Illu- 
mination Measurements." 

By  Dr.  Herbert  E.  Ives. 
"Railway  Car  Lighting." 

By  Air.  Geo.  H.  Hulse. 
"The  Mercury  Quartz  Tube  Lamp." 

By  Mr.  Buckman. 

Friday,  January  6. 
"Deficiencies  of  the  Method  of  Flicker 
for    the    Photometry    of    Lights    of 
Different  Colors." 

By  Prof.  C.  E.  Ferree. 
Saturday,  February  7. 
Meeting  under  the  Auspices  of  Drexel 

Institute. 
"Light  and  How  to  Use  It." 

By  Mr.  C.   O.  Bond,   President 
of  I.  E.  S. 
Wednesday,  February  18. 
Joint  Meeting  with  Franklin  Institute. 
"Artificial  Daylight." 

By  Dr.  Herbert  E.  Ives. 
Friday,  March  20. 
"Lighting    and    Signalling    Systems    of 
Subways." 

By  Mr.  F.  D.  Bartlett. 
"The  Sun— The  Master  Lamp." 

By  Prof.  James  Barnes. 
Thursday,  April  9. 
Joint  Meeting  with  Franklin  Institute. 
"Recent    Developments    in    the    Art    of 
Illumination." 

By  Air.  Preston  S. . Millar. 

Friday,  April  17. 
"The  Structure  of  the  Normal  Eye  and 


its  Ability  to  Protect  Itself  Against 
Ordinary  Light." 

By  Dr.  Wendell  Reber. 
"Glassware  for  Illumination  and  Other 
Purposes." 

By  Mr.  James  Gillinder. 

Friday,  AIay  15. 

Mass    Aleeting   of    all    the    Engineering 

Societies  of  Philadelphia  and 

Vicinity. 

Special  Program  to  be  arranged  and  to 

include  an  address  on 
"The     Relation     of     Engineers    to    the 
Progress  of  Civilization." 

By  Dr.  Chas.  Proteus  Steinmetz. 

PITTSBURGH    SECTION 

A  meeting  of  the  Pittsburgh  Section 
was  held  October  17  in  the  Auditorium 
of  the  Engineering  Society  of  Western 
Pennsylvania.  Sixteen  members  and 
guests  were  present.  Air.  L.  L.  Hopkins 
reviewed  the  proceedings  of  the  Pitts- 
burgh Convention  and  Dr.  H.  H.  Turner 
gave  a  paper  entitled  "The  Essential 
Elements  of  Vision."  Dr.  Turner's 
paper  was  supplemented  with  a  series 
of  slides  and  models. 

The   following  program   of    meetings 
has  been  announced  tentatively: 
November — "Technical     Discussion     of 
the  Elements  of  Lighting"  by  Prof. 
Hower  and  others. 
December — A    joint    meeting    with    the 
Pittsburgh  Section  of  the  Ameritan 
Institute  of  Electrical  Engineers.    A 
Central   Station  paper  will  be  pre- 
sented   by    H.    N.    Muller    of    the 
Duquesne  Light  Company. 
January — A  paper  to  be  selected  by  the 
members  from  Cleveland.    The  sub- 
ject will  be  announced  later. 
February — "Railroad  Yard  Lighting"  by 
A.  C.  Cotton  and  A.  Kirschberg  of 
the     Pennsylvania     Railroad     Com- 
pany. 


TRANSACTIONS    I.    K.    S.— PART    I 


March — A  gas  lighting  subject;  the 
speaker  to  be  announced  later. 

April — "Developments  of  Flame  Carbon 
Arc  Lamps"  by  C.  E.  Stephens. 

May — "Store  Lighting" ;  speaker  to  be 
announced  later. 

June — Open. 


New  Members. 

The  following  twelve  applicants  were 
elected  members  of  the  society  at  a 
meeting  of  the  Council,  November  14, 
I9I3- 

De  Vine.  H.  C. 

Manager,  Pittsburgh  Lamp,  Brass  & 
Glass  Company,  731  Arch  Street, 
Philadelphia,  Pa. 

Fitch,  W.  S. 

Construction  Engineer,  Dennison 
Mfg.  Company,  Framingham,  Mass. 

Hageman,  Jacques  R.  G. 

Engineering  Department,  Bell  Tele- 
phone Company,  2129  Ritner  Street, 
Philadelphia,  Pa. 

Hass,  Henry  P. 

Chief  Inspector,  Department  of 
Tests,  N.  Y.  N.  H.  &  H.  R.  R.  Com- 
pany, New  Haven,  Conn. 

Hicks,  Leslie  R. 

Superintendent,  Fall  River  Electric 
Light  Company,  14  Bedford  Street, 
Fall  River,  Mass. 

Hostetter,  John  S. 

Manager  Fixture  Department,  Bar- 
den  Electric  &  Machinery  Company, 
in  Main  Street,  Houston,  Tex. 

Johnston,  R.  J. 

Testing  Department,  General  Elec- 
tric Company,  Schenectady,  N.  Y. 

Jordan,  Horace  W. 

Illuminating  Engineer,  Edison  Elec- 
tric Illuminating  Company  of  Bos- 
ton, 39  Boylston  Street,  Boston,  Mass. 


Mercer,  J.  M. 

Car  Lighting  Engineer,  The  Adams 
&  Westlake  Company,  319  West 
Ontario  Street,  Chicago,  111. 

Osborn,  Frederick  A. 

Professor  of  Physics,  University  of 
Washington,  Seattle,  Wash. 

Staley,  V.  L. 

General  Electric  Company,  Harri- 
son, N.  J. 

Taylor,  Frank  C. 

Assistant  in  Electrical  Engineering 
Department,  Rochester  Railway  & 
Light  Company,  Rochester,  N.  Y. 


Additional  Sustaining  Members. 

The  following  organizations  were  re- 
cently elected  sustaining  members: 
Gill  Brothers  Company. 

101  Park  Avenue,  New  York,  N.  Y. 

Official  Representative:    John  Beis- 

wanger. 
Jefferson  Glass  Compa:  y. 

Follansbee,  W.  Va. 

KOERTING  &  MATHEISEN. 

22  East  21st  Street,  New  York,  N.  Y. 
Official     Representative:        Charles 
Arnold  Chapin. 
The  Electric  Power  Company,  Ltd. 
506    Confederation    Life    Building, 
Toronto,  Can. 

Official  Representative :  Wills  Mach- 
bachlan. 


Personal. 


Mr.  C.  O.  Bond,  president  of  the 
Illuminating  Engineering  Society,  was 
recently  awarded  the  Beal  medal  of  the 
American  Gas  Institute  for  the  best 
paper  read  at  the  1912  Convention  of 
that  organization.  Mr.  Bond's  paper, 
entitled  "Photometry  of  Incandescent 
Gas  Lamps,"  appears  in  the  1913  Pro- 
ceedings of  the  Institute. 


TRANSACTIONS   I.    E.    S. — PART   I 


Mr.  W.  H.  Gartley,  a  past  president 
of  the  Illuminating  Engineering  Society, 
was  recently  elected  President  of  the 
American  Gas  Institute. 

Hollis  Godfrey,  Ph.  B.,  Sc.  D.,  F.  R. 
G.  S.,  was  recently  elected  president  of 
the  Drexel  Institute  of  Art,  Science  and 
Industry,  Philadelphia,  by  the  trustees 
of  that  institution.  He  will  assume  the 
presidency  December  I,  1913.  Dr.  God- 
frey is  well  known  as  an  educator,  a 
business  man  and  engineer.  He  organ- 
ized the  department  of  science  in  the 
High  School  of  Practical  Arts  in  Bos- 
ton, and  for  four  years  served  as  its 
head.  He  also  spent  six  years  in  night 
school  work  in  Boston  along  the  same 
lines  which  the  Drexel  Institute  is  being 
conducted.  With  two  other  leaders  in 
education  he  organized  the  Garfield 
school,  and  for  two  years  directed  its 
policy  in  the  teaching  of  science,  and 
in  extension  work.  For  three  years  he 
served  as  a  member  of  the  Board  of 
Visitors  of  Tufts  College,  and  was  a 
member  of  the  alumni  council  of  the 
Massachusetts  Institute  of  Technology 
for  two  years.  He  has  outlined  a  course 
in  industrial  engineering  for  the  Society 
of  Promotion  of  Engineering  Education, 
and  has  been  a  lecturer  and  consultant 
in  a  number  of  educational  institutions, 
among  which  are  Dartmouth  College, 
Simmons  College  and  the  University  of 
Wisconsin.  For  several  years  past  he 
has  been  chief  of  the  Bureau  of  Gas  of 
Philadelphia,  and  recently  devoted  an 
exhaustive  study  to  the  organization  and 
operation  of  the  water  bureau  of  the 
latter  city.  In  addition  to  outlining  a 
new  lighting  plan  for  Atlantic  City,  he 
has  also  been  a  consultant  in  the  restora- 
tion of  the  lighting  of  Independence 
Square,  Philadelphia.  He  is  the  author 
of  a  book  on  sanitary  engineering  en- 
titled  "The   Health    of    the    City,"    two 


books  on  "Chemistry,"  and  a  number 
of  monographs  on  scientific  subjects. 
Besides  having  membership  in  a  num- 
ber of  clubs  Dr.  Godfrey  is  a  member 
of  the  Phi  Beta  Kappa  and  the  Theta 
Delta  Chi  fraternities,  the  American 
Society  of  Mechanical  Engineers,  the 
American  Public  Health  Society  and 
the   Illuminating   Engineering   Society. 

Obituary. 

George  H.  Hoffman,  a  district  man- 
ager of  the  Philadelphia  Electric  Com- 
pany, died  in  Philadelphia,  November  3, 
after  a  brief  illness.  He  was  born  in 
New  York  City,  December  22,  1848,  and 
received  a  public  school  education. 
From  1873  to  1877  Mr.  Hoffman  was  a 
member  of  the  New  York  School 
Board,  and  from  1882  to  1889  was  a 
member  of  the  City  Council.  He  was 
appointed  by  President  Cleveland  As- 
sistant United  States  Appraiser  in  1885, 
and  served  in  this  position  until  1889. 
Later  he  engaged  in  the  wool  business 
and  was  also  connected  with  the  Nord- 
Amerika,  a  German  newspaper.  In 
June,  1895,  he  identified  himself  with 
the  electric  lighting  industry  as  mana- 
ger of  the  West  End  Electric  Light 
Company  of  Philadelphia.  In  1901, 
when  the  West  End,  Columbia, 
Diamond  and  Wissahickon  companies 
were  consolidated  into  the  Philadelphia 
Electric  Company,  he  was  made  man- 
ager of  the  northwestern  district  of  the 
company.  He  was  a  charter  member 
of  the  Pen  and  Pencil  Club  of  Phila- 
delphia, and  had  served  as  vice-president 
of  the  International  League  of  Press 
Clubs.  He  took  an  active  interest  in  the 
affairs  of  the  Philadelphia  Section  of 
the  Illuminating  Engineering  Society 
besides  being  a  member  of  several 
other  societies. 


TRANSACTIONS 

OF  THE 

Illuminating 
Engineering  Society 

NOVEMBER,    1913 
PART  II 

Papers,  Discussions  and  Reports 


[  NOVEMBER,  1913  ] 

CONTENTS  -  PART  II 


The  Cooling  Effect  of  Leading-in  Wires  upon  trie  Filaments 
of  Tungsten  Incandescent  Lamps  of  the  Street  Series 
Type.     By  T.  H.  Amrine 385 

Some  Theoretical  Considerations  of  Light  Production.  By 
W.  A.  Darrah 400 

The  Pentane  Lamp  as  a  Working  Standard  By  E.  C. 
Crittenden  and  A.  H.  Taylor 410 

Experiments  in  the  Illumination  of  a  Sunday-School  Room 
with  Gas.     By  Edwin  F.  Kingsbury 439 

Characteristics  of  Enclosing  Glassware.  By  Van  Rensselaer 
Lansingh    447 

The  Photo-Electric  Cell  in  Photometry.  By  F.  K.  Richt- 
myer 459 

Factory  Lighting.     By  M.  H.  Flexner  and  A.  O.  Dicker  • .  470 

Hospital  Lighting.     By  William  S.  Kilmer 488 

Store  Lighting.     By  J.  E.  Philbrick 499 

Distinctive  Store  Lighting.  By  Clarence  L.  Law  and  A.  L. 
Powell 515 

Recent  Improvements  in  Incandescent  Lamp  Manufacture. 
By  Ward  Harrison  and  Evan  J.  Edwards 533 


V4* 


/ 


THE  COOLING  EFFECT  OF  LEADING-IN  WIRES  UPON 

THE  FILAMENTS  OF  TUNGSTEN  INCANDESCENT 

LAMPS  OF  THE  STREET  SERIES  TYPE.* 


BY  T.   H.   AMRINE. 


Synopsis:  This  study  of  the  cooling  effect  of  the  leads  upon  the 
filaments  of  street  series  lamps  was  undertaken  as  a  part  of  a  general 
investigation  into  the  effect  that  dimensions  and  material  of  lead  wires 
and  supports  have  upon  incandescent  lamp  design.  The  method  used  was 
to  calculate  the  average  per  cent,  of  normal  candle-power  and  average 
per  cent,  of  normal  wattage  over  the  cooled  portion  of  the  filament  by 
means  of  measurements  made  upon  two  sets  of  lamps  of  exactly  the  same 
construction  except  having  different  filament  lengths.  The  variation  of 
the  cooling  effect  with  lead  material,  lead  diameter,  lead  length,  filament 
diameter  and  filament  material  was  determined.  The  cooling  effect  of 
any  lead  upon  any  filament  was  shown  to  be  dependent  mainly  upon 
(i)  the  resistance  to  heat  flow  presented  by  the  leads  and  the  cooled  por- 
tion of  the  filament,  (2)  the  diameter  of  the  filament  and  (3)  the  maxi- 
mum temperature  of  the  filament,  i.  e.,  the  temperature  of  the  uncooled 
portion.  For  lamps  having  lead  and  filament  dimensions  encountered  in 
street  series  lamps  the  cooling  effect  decreases  with  increase  of  lead 
length,  with  decrease  in  lead  diameter  and  with  increase  in  thermal 
resistance  of  the  material  of  the  lead.  With  lamps  having  the  same  lead 
construction  but  with  filaments  of  different  diameters  the  cooling  effect 
shows  a  maximum  at  a  value  of  filament  diameter  which  is  dependent 
upon  the  lead  construction  used. 

The  accuracy  with  which  it  is  possible  to  predetermine  the 
dimensions  of  the  filament  of  a  lamp  having  the  desired  candle- 
power,  wattage  and  construction  is  at  present  very  largely 
limited  by  the  lack  of  knowledge  of  the  amount  of  cooling 
effect  of  lead  wires  and  supporting  anchors.  A  length  of  drawn 
tungsten  wire  of  a  certain  diameter  and  carrying  a  certain  cur- 
rent will,  when  subject  to  no  cooling  effect,  consume  a  definite 
number  of  watts  per  centimeter  length  and  will  produce  an 
equally  exact,  if  not  so  readily  determined,  candle-power  per  unit 
length.     The  determination  of  the  total  wattage  and  total  flux 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


386  TRANSACTIONS  I.  E.   S. — PART  II 

of  light  from  such  a  filament  of  any  length  would,  therefore, 
involve  only  multiplication  and  a  correction  for  the  light  absorbed 
by  the  bulb.  The  presence  of  leading-in  wires  and  anchors  sub- 
jects the  filament  to  a  cooling  effect  which  is  dependent  upon 
their  material  and  dimensions  and  upon  the  number  of  anchors. 
This  cooling  effect  makes  the  average  watts  per  centimeter  and 
candle-power  per  centimeter  of  a  given  filament  dependent  not 
solely  upon  the  dimensions  of  the  filament  and  the  current  flow- 
ing through  it,  but  also  upon  the  construction  of  the  lamp  as 
regards  leads  and  anchors. 

The  work  reported  in  this  paper  was  undertaken  as  a  part 
of  a  general  investigation  into  the  effect  that  dimensions  and 
material  of  lead  wires  and  supports  have  upon  incandescent 
lamp  design.  The  study  of  the  cooling  effect  of  the  leads  in  high 
current,  low-voltage  lamps,  such  as  street  series  lamps  was  taken 
up  first  on  account  of  the  fact  that  the  leads  of  these  lamps  cause 
heat  conduction  losses  that  are  large  and  relatively  easily  meas- 
ured. An  investigation  into  the  effect  which  changes  in  such 
variables  as  lead  diameter,  lead  length,  material  of  lead,  filament 
diameter,  filament  length  and  filament  temperature,  can,  there- 
fore, be  carried  on  with  such  lamps  and  data  obtained  which  is  of 
value  not  only  in  connection  with  the  design  of  lamps  of  this 
type,  but  also  in  that  it  will  furnish  some  information  as  to  the 
general  laws  governing  the  relation  between  cooling  effect  and 
the  variables  mentioned.  This  information  as  to  the  approxi- 
mate nature  of  these  laws  will  serve  as  a  guide  in  the  investiga- 
tions upon  the  ordinary  multiple  lamp  where  the  cooling  effect  is 
less  marked  and  less  easily  measured. 

Hyde,  Cady  and  Worthing  have  published*  the  results  of  an 
investigation  into  the  energy  losses  in  lamps  of  the  multiple  type 
and  have  shown  the  variation  in  the  conduction  losses  due  to 
change  in  filament  material  and  filament  temperature.  The 
variations  due  to  lead  dimensions  and  material  and  filament  di- 
ameters were  not  determined  in  this  investigation,  and  no  data 
seems  to  have  been  published  on  this  phase  of  the  subject. 

*  A  Study  of  The  Energy  Losses  in  Electric  Incandescent  Lamps.  E.  P.  Hyde,  F.  E- 
Cady,  A.  G.  Worthing.     Trans.  I.  E.  S.,  Vol.  VI.,  No.  4,  page  238. 


AMRINE:     C00UNG  EFFECT  OF  LEADING-IN  WIRES  387 

METHODS  OF  INVESTIGATION. 

A  number  of  methods  of  carrying  out  this  investigation  are 
possible  and  were  considered  before  the  work  was  taken  up. 
A  strictly  mathematical  method  can,  of  course,  be  employed,  the 
temperature  of  the  filament  at  different  distances  from  the 
leads  being  calculated  and  the  total  changes  in  candle-power  and 
wattage  due  to  the  cooling  effect  determined  therefrom.  On 
account  of  the  number  of  variables  and  complexity  of  the  prob- 
lem, the  mathematical  method  yields  a  rather  unwieldy  equation. 
Its  accuracy  is  dependent  upon  the  thermal  constants  of  tungsten 
at  high  temperatures  and  upon  the  law  of  their  change  with  the 
temperature  and  upon  the  thermal  resistance  of  the  welded 
joints  between  the  lead  wire  and  filament  and  other  factors  upon 
which  there  is  no  good  data.  A  much  preferable  method  is  to 
measure  directly  the  temperature  of  the  filament  at  points  at 
different  distances  from  the  leads  by  some  means  such  as  a 
thermocouple  made  of  very  fine  wires  or  by  an  optical  pyrometer 
such  as  was  used  by  Hyde,  Cady  and  Worthing.  These  tempera- 
ture measurements  are  laborious  to  make  by  either  method  and 
the  time  required  to  cover  the  ground  which  was  desired  to  cover 
in  this  investigation  would  be  very  great.  By  the  optical  pyrom- 
eter method  measurements  of  temperature  over  the  first  two 
or  three  millimeters  of  the  filament  adjacent  to  the  lead  of  a 
lamp  of  the  series  type  are  very  unsatisfactory  on  account  of  the 
very  low  luminosity  of  the  filament  at  these  points.  For  these 
reasons  the  methods  involving  temperature  measurements  were 
not  adopted,  though  for  the  purposes  of  studying  the  temperature 
conditions  adjacent  to  the  leads  a  considerable  number  of  tests 
were  made  by  a  method  only  slightly  modified  from  that  used 
by  Hyde,  Cady  and  Worthing.  The  modification  was  simply 
the  interchanging  the  positions  of  the  lamp  under  test  and  the 
comparison  lamp  so  as  to  permit  of  the  latter  burning  at  a  lower 
temperature. 

The  method  adopted  was  to  arrive  at  the  figure  for  the 
average  per  cent,  of  normal  wattage  and  average  per  cent,  of 
normal  candle-power  over  the  cooled  portion  of  the  filament 
by  means  of  measurements  made  upon  two  sets  of  lamps  having 
different  lengths  of  filaments.     This  method  was  made  possible 


388  TRANSACTIONS  I.  E.  S. — PART  II 

by  the  fact  that  modern  tungsten  wire  drawing  methods  permit 
of  the  manufacture  of  wire  of  almost  exactly  uniform  diameter 
and  composition  throughout  its  length.  The  average  candle- 
power  and  wattage  of  two  sets  of  lamps  of  exactly  the  same  con- 
struction can,  therefore,  be  depended  upon  to  be  very  nearly  the 
same.  Also  the  difference  between  the  average  candle-power, 
for  instance,  of  a  set  of  lamps  of  a  certain  construction  and  that 
of  another  set  of  exactly  the  same  construction,  except  for  fila- 
ment length,  will  be  the  candle-power  of  an  uncooled  portion  of 
the  filament  equal  in  length  to  the  difference  between  the  average 
filament  lengths  of  the  two  lots  of  lamps. 

Assume  that  it  is  desired  to  determine  the  cooling  effect  upon 
a  filament  of  a  given  diameter  of  a  lead  of  given  dimensions 
and  material.  Two  lamps  are  made  from  wire  of  the  correct 
diameter  taken  from  the  same  spool  and  having  the  desired 
dimensions  and  material  of  leads  and  of  exactly  the  same  con- 
struction throughout,  except  that  one  has  a  filament  of  length  1 
and  the  other  a  filament  of  length  1'.  Let  c  and  w  represent 
the  candle-power  per  centimeter  length  and  the  watts  per  centi- 
meter length  respectively  of  the  uncooled  portion  of  the  filament 
with  a  current  passing  through  the  lamp  which  will  bring  this 
uncooled  portion  of  the  filament  to  the  desired  temperature — 
say  that  corresponding  to  normal  operation.  Let  a  equal  the 
total  length  of  the  cooled  portion  of  the  filament  in  each  case. 
This  will  be  the  same  for  two  lamps  of  similar  construction  ex- 
cept for  filament  length,  if  the  filament  is  longer  than  a  certain 
minimum  length.  The  expressions  for  the  total  candle-power 
and  total  watts  of  the  two  lamps  will  be 

C=aK-|-(l — a)c,    for   lamp   of   filament   length   1 
W=aP-f-(l — a)w,   for  lamp  of  filament  length  1 
Similarly, 

C'^aK-l-O' — a)c,  for  lamp  of  filament  length  1' 
W'=aP+0' — a)w,  for  lamp  of  filament  length  1' 
where  K  is  the  average  per  cent,  candle-power  and  P  the  average 
per  cent,  wattage  over  a  cooled  portion  of  the  filament  a  centi- 
meters long.  These  figures  for  average  per  cent,  are  based  upon 
the  candle-power  and  wattage  of  a  centimeters  of  the  uncooled 
portion  of  the  filament. 


AMRINE  :     COOLING  EFFECT  OF  LEADING-IN  WIRES  389 

C  and  C  can  be  determined  by  careful  photometer  measure- 
ments and  W  and  W  by  means  of  a  potentiometer.  An  investi- 
gation with  an  optical  pyrometer  showed  that  in  none  of  the 
lamp  constructions  which  it  was  planned  to  investigate  was  there 
any  cooling  effect  at  a  distance  greater  than  two  centimeters.  For 
the  sake  of  uniformity  this  distance  of  2  centimeters  was  adopted 
arbitrarily  as  the  length  of  the  cooled  portion  adjacent  to 
one  lead.  Hence  a  is  equal  to  4  centimeters.  Between  the  above 
equations,  K  and  P  and  the  values  of  c  and  w  can  be  determined, 
since  1,  1',  C,  C,  W,  W  and  a  are  known.  The  values  of  K  and 
P  based  on  the  standard  distances  of  2  centimeters  from  each 
lead,  are  measures  of  the  cooling  effect.  By  using  these  values 
and  the  values  of  c  and  w,  as  determined  above,  one  can  calculate 
the  candle-power  and  watts  of  an  ideal  filament  {i.e.,  one  which 
is  not  subject  to  a  cooling  effect)  of  length  1  or  1'  and  compare 
the  values  obtained  with  C,  C,  and  W  and  W. 

A  quantity  M  was  also  calculated  for  each  case  studied.  This 
quantity  is  the  per  cent,  difference  between  the  wattage  of  an 
ideal  lamp  with  an  uncooled  filament  of  10  centimeters  length 
and  that  of  a  lamp  of  any  given  lead  construction  of  the  same 
candle-power  and  having  a  filament  of  the  same  diameter  and 
operating  at  the  same  current.  M  expresses  by  one  quantity  the 
result  of  cooling  effect  on  both  the  candle-power  and  watts  of  a 
filament. 

This  method  has  the  advantage  that  the  work  required  for 
one  determination  can  be  carried  out  in  a  reasonable  length  of 
time  so  that  the  various  experiments  can  be  made  upon  a  suffi- 
cient number  of  lamps  so  that  a  good  average  result  can  be  ob- 
tained. In  this  work  on  an  average  of  four  lamps  of  each  con- 
struction were  investigated  so  that  it  is  felt  that  fairly  good 
average  values  were  obtained. 


EXPERIMENTAL  RESULTS. 
The  lamps  used  in  the  experiments  were  similar  in  construction 
to  the  regular  street  series  tungsten  lamps  except  that  there  were 
no  anchors  or  supports  used.  All  lamps  had  drawn  wire  tungsten 
filaments  and  were  made  in  straight  sided  bulbs  of  4^  inches 
(11. 11   centimeters)    maximum   diameter    (commercially  known 


39°  TRANSACTIONS  I.   E.   S. — PART  II 

as  S-35  bulb)  and  with  a  skirted  screw  base  (commercially  known 
as  No.  108  base)  and  with  the  same  size  of  glass  stems.  The 
dimensions  of  the  lead  wires,  of  course,  were  different  in  the 
different  experiments.  Fig.  1  shows  the  type  of  lamps  experi- 
mented upon. 

The  routine  followed  in  carrying  out  these  experiments  was  as 
follows :  Assuming,  for  example,  it  was  desired  to  compare  the 
cooling  effect  on  a  filament  of  a  given  diameter  and  tempera- 
ture, of  leads  of  two  different  diameters.  Ten  lamps  were  made 
up  with  filaments  taken  from  the  same  sample  of  wire  with 
each  of  the  desired  lead  constructions,  all  lamps  of  each  lot 
of  ten  being  exactly  alike  except  that  five  had  filaments  of  about 
6  centimeters  length  and  five  had  filaments  of  about  15.5  centi- 
meters length.  After  the  filaments  were  mounted  they  were 
each  carefullly  measured  before  sealing-in  for  filament  length 
between  leads.  The  lamps  were  exhausted  in  the  regular  manner, 
based  and  then  all  carefully  measured,  at  the  amperes  to  give 
the  desired  temperature,  for  candle-power  upon  a  precision 
photometer  and  for  wattage  by  means  of  a  potentiometer.  The 
values  for  K,  the  average  per  cent,  of  normal  candle-power, 
and  P,  the  average  per  cent,  of  normal  wattage  each  over  a 
portion  of  filament  to  a  distance  of  2  centimeters  from  each  lead, 
were  then  calculated  as  shown  above.  From  these  values  the 
value  of  M,  the  per  cent,  wattage  change  due  to  cooling,  were 
calculated.  The  average  figures  obtained  with  each  lot  of  lamps 
were  used  as  the  correct  values  of  K,  P  and  M. 

On  account  of  the  fact  that  the  accuracy  of  the  wattage  meas- 
urements is  considerably  better  than  that  of  the  candle-power 
measurements,  the  curves  shown  were  based  upon  the  wattage 
measurements.  From  the  relation  between  K  and  P  determined 
from  all  the  experimental  data,  the  points  on  the  curves  for  K 
as  plotted  were  calculated  from  the  values  of  P.  In  this  way 
the  K  curves  were  "smoothed  out"  and  made  to  conform  with  the 
P  curves.  By  this  plan  it  was  thought  that  the  relation  between 
K  and  different  variables  as  lead  length,  lead  diameter,  etc.,  are 
more  accurately  shown  and  that  the  absolute  values  of  K,  as 
plotted,  are  probably  as  accurate  as  the  individual  experimental 
values  obtained. 


3qO 


^ 

**■ 

t. 

yv 

T  jl^ 

I 

Fig.  i. 


M      K 
1 1.0    62 


56 
10.0    54 

52 
9.5     50 

48 
9.0   46 


\M 

^ 

,j 

""    "s. 

K  . 

^^^ 

■'* 

T    ~ 

CURVES  SHOWING  COOLING  EFFECT  OF 
LEADSOF  DIFFERENT  LENGTHS  UPON 

A  FILAMENT  .01065  DIAM.  CARRYING  fc5 

AMPjS.  LEADS  OF  COPPER  .05;DIAM. 

LENGTH    IN    INCHES 


62 
10.5   60 


54 
a5   52 


^\p 

p 

>~^ 

^^_ 

K 

~~. 

--.^ 

^><C 

Hr+N 

M, 

*■»«! 

^ 

s^\ 

CURVES  SHOWING  COOLING  EFFECT  OF  LEADS  Oc  "^^ 
DIFFERENT  D  ivE-F-=s  .":'.  A  r  _A".'E\T.OIQ6r>    \ 

DIAMETER  CARRYING  6.5  AVP  LEAD:  C^COP-E3  :=  _:'.;.  > 

> 

81 

LEAD  DIAMETER 

Fig.  3- 


A 

r 

^\ 

1/ 

/K 

frv 

:jr. 

0F.0 

ES  SHOWING  COOLIN 
y«8 CU  LEADS  UPON  Fl 

jEFF 
-AME 

-A 

i-s\ 

X 

• 

^». 



— s? 

,^ 

.003         .010  .Oil         .012  .013 

FILAMENT  DIAMETER  IN  INCHE5 

Fig.  4. 


AMRINE  :     COOLING  EFFECT  OF  LEADING-IN  WIRES  391 

Fig.  2  shows  the  values  obtained  at  operating  efficiency  for 
K,  P  and  M  for  lamps  having  different  lengths  of  leads.  In 
these  and  in  the  following  curves  K  is  the  average  per  cent, 
candle-power  and  P  the  average  per  cent,  wattage  over  the  two 
centimeters  of  the  filament  adjacent  to  each  lead;  that  is,  the 
section  of  the  filament  which  embraces  the  cooled  portion.  These 
values  are  based  upon  the  candle-power  and  wattage  of  two  cen- 
timeters of  a  similar  but  uncooled  filament  carrying  the  same 
current.  M  is  the  per  cent,  difference  in  wattage  between  an 
ideal  filament  and  a  cooled  filament  as  defined  under  the  heading 
of  Method  of  Investigation.  These  curves  indicate  that  for  the 
lead  construction  used  the  longer  the  lead,  the  less  the  cooling 
effect  as  is  shown  by  the  higher  values  of  K  and  P  and  the  lower 
values  of  M.  The  lead  length  as  used  in  this  connection  is  taken 
as  the  length  of  lead  from  filament  to  the  glass  stem. 

Fig.  3  shows  the  greater  cooling  effect  of  large  diameter  leads 
as  compared  with  small  diameter  leads  of  the  same  material. 

In  order  to  compare  the  relative  cooling  effects  of  copper  and 
nickel  leads  of  the  same  dimensions,  K,  P  and  M  were  deter- 
mined for  a  6.5  ampere  filament  using  leads  0.050  in.  in  diameter 
and  iy2  in.  (3.81  cm.)  long  in  each  case.  The  values  obtained 
were  as  follows : 

p  K  M 

Copper 82.6  56.3  10.52 

Nickle 90.6  67.5  9.24 

The  influence  of  filament  diameter  upon  cooling  effect  is  shown 
by  the  curves  of  Fig.  4.  The  rather  unexpected  shape  which 
these  curves  take  is  no  doubt  accounted  for  by  the  fact  that  the 
total  resistance  to  heat  flow  from  the  filament  is  made  up  of 
three  parts,  that  due  to  the  leads,  that  due  to  the  welded  joints  and 
that  due  to  the  cooled  portion  of  the  filament.  With  the  same 
size  leads  and  the  same  sort  of  a  welded  joint  this  total  resistance 
might  be  expected  to  show  a  minimum  at  some  value  of  filament 
diameter.  In  a  lamp  of  the  construction  studied  this  minimum 
occurs  at  about  0.0108  in.  diameter. 

Fig.  5  is  a  set  of  curves  which  shows  the  variation  in  cooling 
effect  with  the  filament  temperature.      These  curves  are  plotted 

between  "per  cent,  of  normal  -= — -"  and  "percent,  of  normal  P," 

^3/2 


392 


TRANSACTIONS  I.  E.   S. — PART  II 


"per  cent,  of  normal  K  and  "per  cent,  of  normal  M."  The 
quantity  ^— r  is  a  measure  of  filament  temperature,  I  represent- 
ing the  current  through  the  filament  and  d  the  diameter  of  the 
filament.  A  value  of  -=— p  equal  to  0.0059  was  taken  as  normal. 
This  corresponds  to  approximately  1  watt  per  projected  candle- 


110 

M 

1 

95 

P. 

V 

90 

K/ 

CURVES  SHOWING  VARIATION    Op  %  OF  NORMAL  P.K  &M  Vl 
ViS  BASED  ON  A  VALUE  OF  Virt  CORRESPONDING  TO  i.oc 
WATT/PROJECTED  CP  AS  NORMAL. 

ITU 

|       | 

'/d*  *  1000 
Fig.  5- 


220O 
2100 

2000 

)  1900 

c 

;  1800 

\  1700 

2  isoo 

J 

51500 

J 

i|400 

1300 


■ 

| 

k. 

3.2  AMR 

7.5  AMR 

1 

V 

6.5  AMP 

5.5  A 

■;p 

CURVES  SHOWING-  TEMPERAIUKtb  IN 
DES.C.  OF  A  FILAMENT  AT  VARIOUS  AMPS. 
AND  VARYING  DISTANGE5FR0M  THE  LEADS._ 

7.5A^ 

RNOf 

MAL 

06V  l| 

OJ.L 

""""O      2       4        6       8      10      12      W      16      18      20     22      24     26 

DISTANCES  FROM  LEAD  IN  MM. 

Fig.  6. 

power.  It  also  corresponds  approximately  to  the  filament  tem- 
peratures used  in  the  measurements  made  upon  the  relative  cool- 
ing effects  of  leads  of  different  diameters,  lengths  and  material. 
These  curves  indicate  that  with  higher  filament  temperatures, 
i.  e.,  lower  watts  per  candle-power,  the  cooling  effect  is  less 


AMRINF  :     COOLING  EFFECT  OF  LFADING-IN  WIRES 


393 


marked.  This  can  be  readily  noticed  in  a  lamp  if  the  cooled  por- 
tion of  a  filament  is  observed  first  with  a  large  current  flowing 
and  then  with  a  smaller  one.  In  the  latter  case  the  filament  is 
cooled  out  to  a  considerably  greater  distance  from  the  lead.  The 
curves  of  Fig.  6,  the  data  of  which  was  obtained  by  the  optical 
pyrometer  method,  show  the  decrease  in  temperature  near  the 
leads  and  the  greater  cooling  effect  at  higher  temperatures. 

A  study  of  the  foregoing  data  shows  that  the  cooling  effect  of 
any  lead  upon  any  filament  is  dependent  mainly  upon  three  things : 
(i)  the  resistance  to  heat  flow  presented  by  the  leads  and  the 
cooled  portion  of  the  filament,  (2)  the  diameter  of  the  filament 
and   (3)   the  maximum  temperature  of  the  filament,  i.  e.,  the 


Kooo 


12000 


10.000 


fcooo 


CURVES  SHOWING  RELATION  BETV 
TOTAL  THERMAL  RESISTANCE (LT.J36F 
L" THERMAL  RESISTANCE  OF  LEAD. 

'EEN_ 
AHD|. 

TF-      -                 "      OFT"  FILAM 

RELATIVE  THERMAL  C0NDUCTIV1T1E 
1      CU.-5.07.  NICKEL-.83G.  TUNCSTEN-2 

:NT. 
8- 

2i 

O 

<> 

tOO  500 

I. +.256  F 


Fig.  7- 

temperature  of  the  uncooled  portion.  With  a  given  maximum 
filament  temperature  and  a  given  thermal  resistance  due  to  the 
leads  and  the  cooled  portion  of  the  filament  the  same  number  of 
watts  will  be  lost  by  conduction  (neglecting  differences  due  to 
the  amount  of  heat  radiated  by  the  lead  and  cooled  portion  of 
the  filament).  With  a  given  number  of  watts  escaping  by  con- 
duction there  will,  of  course,  be  a  greater  per  cent,  reduction  in 
wattage  with  a  filament  of  small  diameter  than  one  with  a  fila- 
ment of  a  larger  diameter.  By  plotting  the  data  obtained  with 
"total  thermal  resistance,"  that  is,  the  resistance  of  the  lead  plus 
that  due  to  the  cooled  portion  of  the  filament,  as  abscissae  and 


—  as  ordinates,  where  d  is  filament  diameter,   an  approximately 


394 


TRANSACTIONS  I.  E.   S. — PART  II 


smooth  curve  is  obtained.  In  this  curve,  Fig.  7,  the  "total  thermal 
resistance"  is  taken  to  include  0.236  in.  of  the  filament  and  is 
calculated  using  lengths  in  inches,  diameters  in  inches  and  the 
relative  thermal  conductivity  of  the  materials  as  follows : 

Copper 5.07 

Tungsten 2.26 

Nickel 0.876 

The  curve  includes  the  data  taken  upon  leads  of  different 
diameters,  lengths  and  materials  and  different  filament  diameters 
and  should  show,  over  the  range  covered  by  the  experiments, 
approximately  the  cooling  effect  of  any  lead  upon  a  filament  of 
any  diameter.  It  does  not,  however,  take  into  consideration  the 
influence  of  change  in  thermal  conductivity  with  temperature, 


^88 

!- 

z 
< 

L 

f 

$&- 

40 

4 

4 

f 

,f 

9 

i* 

j=  30 

r  88 

CURVE 
20C-E- 

SSHO 
SERIES 

viks'h 

JMPH 

EATINf 
TH.O? 

U/OF40AMR 
ijcu  LEADS. 

Fig.  8. 


effect  of  radiation  from  the  leads,  etc.    The  filament  temperatures 

correspond  to  -=— 7-  equal  to  0.0059,   tnat  is.  approximately  one 

watt  per  projected  candle-power.  By  the  use  of  this  curve  and 
the  P  curve  of  Fig.  5,  one  can  determine  approximately  the 
values  of  P  for  filaments  at  other  temperatures.  Knowing  that 
K  varies  about  as  the  cube  of  P  the  value  of  the  average  per  cent, 
candle-power  over  the  cooled  portion  of  a  filament  with  any  lead 
construction  can  be  determined. 

The  data  obtained  in  the  experiments  described  show  that  there 
is  a  marked  cooling  effect  -in  lamps  of  the  construction  used.  That 
this  cooling  effect  is  of  vastly  more  than  mere  scientific  interest 
in  the  case  of  these  lamps  is,  perhaps,  not  generally  recognized 


AMRINE  :     COOLING  EFFECT  OF  LEADING-IN  WIRES  395 

except  by  lamp  specialists.  However,  in  the  design  and  rating  of 
and  measurements  upon  these  lamps  this  cooling  effect  must  be 
taken  into  practical  consideration. 

In  Fig.  8  are  shown  curves  which  demonstrate  that  the  cooling 
effect  of  leads  must  be  taken  into  consideration  in  all  careful 
measurements  of  series  lamps.  These  curves  were  obtained  by 
holding  the  normal  current  through  the  lamp  while  the  voltage 
and  candle-power  were  measured  at  short  intervals  of  time.  Due 
to  the  cooling  effect  of  the  leads,  the  lamp  does  not  come  up  to 
full  normal  candle-power  until  several  minutes  after  the  current 
is  turned  on.  Errors  that  would  be  quite  serious  in  careful  work 
are  caused  if  series  lamps  are  photometered  or  measured  for 
volts  at  amperes  without  taking  the  precaution  first  to  allow  the 


:'5 

<KL7 


1 

1   1    1    1    1! 

CURVES  SHOWING  AVE?i3F. 

EF*  C_=  V  .»  r.AMEVS  CF  WFTTim 

15  AJ 

■=:. 

UEM 

.:-:'■ 

V 

Fig.  9. 

lamp  to  heat  up  thoroughly.  For  instance,  an  error  of  about 
1.3  candle-power  or  6.5  per  cent,  would  be  caused  in  photometer- 
ing  a  20  candle-power,  4.0  ampere  series  lamp  20  seconds  after 
lighting  up  instead  of  allowing  it  to  heat  up  thoroughly. 

Due  mainly  to  this  influence  of  the  cooling  effect  of  leads,  it 
is  impracticable  to  make,  for  instance,  7.5  street  series  lamps  of 
the  various  candle-power  sizes  at  the  same  efficiency  rating. 
Lamps  of  the  same  current  rating  have,  of  course,  a  candle- 
power  almost  proportional  to  the  filament  length.  The  loss  in 
candle-power  and  the  decrease  in  wattage  for  a  low  candle-power 
lamp  of  a  given  diameter  of  filament  and  a  given  current  flowing 
is  a  much  larger  per  cent,  of  the  total  for  a  short  filament  or  low 
candle-power  lamp  than  with  a  long  filament  or  high  candle- 


39^ 


TRANSACTIONS  I.  E.   S. — PART  II 


power  lamp.  Hence,  if  the  current  in  the  short  filament  lamp  is 
increased  until  the  filament  is  operating  at  the  same  watts  per 
mean  spherical  candle-power  then  its  temperature  out  on  the 
uncooled  portion  of  the  filament  is  much  greater  than  the  maxi- 
mum temperature  of  the  long  filament  lamp  and  will,  therefore, 
give  a  much  shorter  life.  In  Fig.  9  is  a  curve  which  shows  the 
watts  per  mean  spherical  candle-power  of  lamps  made  with  fila- 
ments of  the  same  diameter  and  with  the  same  current  passing 
through  them,  but  of  different  filament  lengths.  It  shows  that 
at  the  same  amperes  a  short  filament  is  operating  at  a  very  much 
poorer  efficiency  than  a  long  filament  and  if  operated  at  the  same 
efficiency  would  give  a  very  poor  life. 


&o 


J07.8 


<ia 


\         =  .-::-      .<,-•!*:  a^e^es  an;  ~ ;_--  . E  '..  .: 
\    |0F  SERIES  LAMPS  HAVIN6  FILAMENTS  OF  DIFFERENT 

HORIZONTAL  CANDLE-PCWE8.                          | 
FILAMENT  .Olia'DIAM.- .OS'*  I5CU.  LEADS 

/r 

<£ 

\ 

_) 

^V/ 

| 

zoo 


'  0      Z      1      <c      8      10     12     H      IS     IS    20    22    24     26    28    30 
CM. LENGTHS 

Fig.  io. 

In  Fig.  io  is  shown  what  would  be  the  effect  if  street  series 
lamps  of  a  given  wire  size  and  current  with  their  great  cooling 
effects  were  rated  at  the  same  watts  per  mean  horizontal  candle- 
power  regardless  of  their  filament  lengths.  These  curves  assume 
7.5  ampere  lamps  of  different  candle-powers,  that  is,  different 
filament  lengths.  They  are  assumed  all  to  be  rated  at  i.i  watts 
per  mean  horizontal  candle-power.  The  curve  marked  "amperes" 
shows  the  amperes  that  it  would  be  necessary  to  pass  through 
the  filaments  of  the  various  lengths  in  order  to  bring  them  to  the 
same  mean  horizontal  efficiency.  The  great  increase  in  amperes 
necessary  to  bring  the  short  filaments  up  to  efficiency,  of  course, 
very  seriously  cuts  down  their  lives,  since  a  lamp's  life  is  deter- 
mined by  the  temperature  of  the  hottest  part  of  the  filament. 
This  decrease  in  life  is  shown  by  the  curve  marked  "life."    The 


AMRINE:     COOLING  EFFECT  OF  LEADING-IN  WIRES 


397 


character  of  these  curves  is  almost  wholly  brought  about  by  the 
cooling  effect  of  the  leads,  though  the  change  in  spherical  reduc- 
tion factor  with  the  filament  length  enters  to  a  minor  extent. 

By  going  to  the  extremes  in  cooling  effect  by  using  filaments 
so  short  that  they  are  cooled  over  practically  their  entire  length, 
one  can  arrive  at  the  unusual  condition  of  having  two  lamps  with 
filaments  of  the  same  diameter  and  differing  in  length  by  30 
per  cent,  or  more  but  with  the  same  voltage  and  candle-power. 
This  is  brought  about  by  the  difference  in  the  volt  candle-power 
characteristic  caused  by  the  difference  in  the  cooling  effect  of  the 
leads  on  the  two  filaments.    Fig.  11  shows  the  volt  candle-power 


A, 

/ 

• 

/ 

f      y 

/a 

£ 

/ 

/  y 

1 
4/ 

vy 

Z 

22 

^ 

* 

:-- 

VOLT  CANDLE-POWER  CHARACTERISI 
6£AMP  LAMPS  wrrH.05V^CU.  LEADS 

cor 

B- 

1.59 

VOLTS 

Fig.  11. 

characteristics  of  two  lamps  made  exactly  alike  except  for  fila- 
ment length.  Lamp  A  had  a  filament  1.31  cm.  long,  lamp  B, 
1.59  cm.  long.  Both  filaments  were  made  of  wire  0.0118  in.  diam- 
eter taken  from  the  same  spool.  The  characteristics  of  the  two 
lamps  intersect  and  show  that  both  lamps  give  2.27  candle-power 
at  1.44  volts.  In  other  words,  if  one  had  a  lamp  of  the  construc- 
tion of  lamp  B,  he  could  remove  17.5  per  cent,  of  the  filament 
and  still  have  a  lamp  of  the  same  candle-power  and  voltage. 

Notwithstanding  the  magnitude  of  the  cooling  effect  on  lamps 
of  the  street  series  type  it  is  for  the  most  part  impracticable  to 
let  this  factor  determine  the  lead  construction  of  these  lamps. 
That  is,  it  is  usually  not  feasible  in  practice  to  so  construct  a 
lamp  that  the  losses  due  to  cooling  effect  approach  a  minimum. 
Other  conditions  will  usually  limit  the  designer's  choice  of  lead 


39§  TRANSACTIONS  I.  E.   S. — PART  II 

dimensions  and  materials.  The  appearance  of  the  lamp,  bulb 
dimensions  and  added  energy  losses  due  to  electrical  resistance 
will  usually  limit  the  length  of  leads.  The  vapor  pressure,  and 
electrical  resistance  of  the  material  of  the  lead  will  usually  deter- 
mine its  diameter  and  these  factors  as  well  as  the  ease  with  which 
the  metal  may  be  subjected  to  the  various  manufacturing  pro- 
cesses will  determine  the  material  of  the  lead.  A  knowledge  of 
the  manner  in  which  the  cooling  effect  varies  with  lead  material 
and  dimensions  is  therefore  chiefly  of  importance  in  that  it 
enables  one  to  make  the  proper  allowances  in  candle-power,  watts 
and  volts  in  designing  a  lamp  to  a  given  specification. 

The  author  wishes  to  acknowledge  his  indebtedness  to  Mr. 
L,.  M.  Moss,  who  assisted  in  all  the  experimental  work  and  in 
the  preparation  of  the  data  for  this  paper. 

DISCUSSION. 

Mr.  Evan  J.  Edwards  :  I  have  consulted  Mr.  Amrine  regard- 
ing a  demonstration  which  was  used  in  the  presentation  of  the 
paper  "Recent  Improvement  in  Incandescent  Lamp  Manufac- 
ture," and  I  believe  it  worth  while  to  repeat  it,  for  it  shows  in  a 
very  striking  manner  the  cooling  of  the  filament  adjacent  to  sup- 
ports and  leading-in  wires.* 

There  is  one  interesting  point  concerning  the  cause  of  the  cool- 
ing which  Mr.  Amrine  has  not  emphasized,  and  that  is  the 
changed  rate  at  which  electrical  energy  appears  as  heat  in  the 
cooled  portions.  The  fundamental  cause  of  the  reduced  tem- 
perature is,  of  course,  the  conduction  of  heat  to  the  supports  or 
leading-in  wires.  But  the  reduced  temperature  condition,  in  the 
case  of  metal  filaments,  brings  about  a  lowering  of  the  resistance 
which  also  contributed  to  lowering  of  the  temperature.  The  cur- 
rent in  all  portions  of  the  filament  is  the  same,  and  therefore, 
the  rate  at  which  heat  is  received  by  any  portion  of  the  filament 
is  proportional  to  the  resistance  of  that  portion. 

Changed  resistance  therefore  has  the  effect  of  helping  to  lower 
the  temperature  of  the  cooled  portion  of  metal  filaments  because 

*  A  stereopticon  was  used  to  project  the  direct  image  of  the  lighted  filament  of  a 
large  special  lamp  on  a  background  lighted  by  the  regular  light  source  of  the  lantern. 
After  adjustments  of  current  in  the  special  lamp  were  made,  the  hottest  parts  of  the  fila- 
ment appeared  as  bright  lines  and  the  cooled  portions  as  dark  lines. 


COOLING  EFFECT  OF  LEADING-IN    WIRES  399 

of  their  positive  resistance-temperature  coefficient.  The  oppo- 
site is  true  for  the  negative  coefficient  carbon  filaments. 

Mr.  M.  Luckiesh:  At  the  bottom  of  the  tenth  page  Mr. 
Amrine  states  "These  curves  indicate  that  with  higher  filament 
temperatures,  i.  e.,  lower  watts  per  candle-power,  the  cooling 
effect  is  less  marked."  This  is  at  once  evident  from  the  funda- 
mental laws  of  radiation.  As  the  temperature  of  the  filament 
increases,  the  conduction  loss  becomes  proportionately  less 
because  it  increases  approximately  directly  with  the  temperature 
while  the  candle-power  is  a  function  of  a  higher  power  of  the 
temperature.  Therefore  as  the  temperature  of  the  filament  is 
increased  the  candle-power  will  increase  much  more  rapidly  than 
the  conduction  losses  in  the  leading-in  wires. 

Mr.  T.  H.  Amrine:  I  wish  to  add  that  the  lamp  with  the 
shorter  filament,  lamp  A,  has  a  wattage  of  11.07  and  lamP  B  has 
a  wattage  of  12.25  at  the  voltage  which  makes  the  candle-powers 
of  the  two  lamps  equal,  that  is,  at  1.44  volts. 


400  TRANSACTIONS  I.   E.   S. — PART  II 

SOME  THEORETICAL  CONSIDERATIONS  OF  LIGHT 
PRODUCTION.* 


BY  W.  A.  DARRAH. 


Synopsis:  This  paper  sets  forth  some  of  the  inherent  limitations  of 
the  various  electric  illuminants  now  in  use,  and  discusses  briefly  the  effect 
of  these  limitations  upon  the  progress  of  the  art  of  illumination.  The 
basic  theory  of  light  production  by  the  acceleration  of  electric  charges 
is  analysed  with  a  view  to  indicating  the  relative  possibilities  of  the 
various  types  of  illuminants  particularly  the  arc  and  the  incandescent 
lamp.  The  structure  of  the  atom  is  considered  in  the  light  of  modern 
theories,  and  some  general  deductions  made  regarding  the  properties 
which  a  substance  is  likely  to  exhibit  if  well  adapted  for  use  as  a  radiating 
body.  A  tentative  explanation  of  the  action  of  selective  radiation  is 
given,  and  of  the  part  which  this  phenomenon  plays  in  present  illuminants. 
It  is  demonstrated  that  to  secure  high  luminous  efficiencies  selective 
radiation  must  be  relied  upon,  and  therefore  unless  new  materials  are 
discovered  which  exhibit  selective  radiation,  while  in  the  solid  state,  the 
efficiency  of  the  electric  arc  will  remain  materially  higher  than  the  efficiency 
of  the  so-called  incandescent  lamp,  where  electrical  energy  is  trans- 
formed into  heat  and  then  into  light  by  means  of  the  resistance  of  a 
conductor. 

It  is  proposed  in  this  paper  to  discuss  some  theoretical  phases 
of  light  production  with  a  view  of  indicating  how  these  consid- 
erations have  influenced  the  progress  of  the  art  of  illumination 
and  their  probable  effect  upon  future  developments  in  the  direc- 
tion of  higher  efficiencies. 

No  claim  is  made  here  for  original  theories,  as  modern  con- 
ceptions and  methods  of  mathematical  analysis  have  been  used 
freely;  but  an  effort  has  been  made  to  apply  these  theories  in  a 
way  to  indicate  the  trend  of  engineering  progress. 

The  generally  accepted,  and  most  satisfactory  conception  of 
light,  is.  that  of  a  transverse  electromagnetic  vibration,  of  the 
ether,  which  travels  in  straight  lines  at  a  speed  of  approximately 
180,000  miles  per  second.  Having  identified  light  as  an  elec- 
tromagnetic vibration  it  is  necessary  to  differentiate  it  from  other 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


DARRAH  :     SOME    CONSIDERATIONS    OF    EIGHT    PRODUCTION    4OI 

electromagnetic  vibrations,  as  the  radiation  utilized  in  wireless 
telegraphy,  the  so-called  "radiant  heat,"  and  the  large  class  of 
radiations  possessing  varying  properties,  and  known  as  x-rays. 
In  present  theories  the  frequency  of  the  vibration  (or  what  is 
practically  its  reciprocal,  the  wave-length)  is  considered  suffi- 
cient to  supply  the  identification,  although  it  seems  probable  to 
the  writer  that  under  some  conditions,  the  wave  form  of  the 
vibration  may  be  as  important  as  the  frequency.  Omitting  this 
point,  the  zone  covered  by  visible  electromagnetic  radiations  is 
comparatively  narrow,  comprising  not  more  than  one  octave  out 
of  45  of  the  entire  section  between  the  longest  electric  wave,  and 
the  shortest,  ultra-violet  wave.  In  other  words,  only  those  vibra- 
tions which  fall  between  400  X  1012  and  800  X  1012  cycles  per 
second  are  visible,  assuming  the  waves  to  be  simple  harmonics. 

Whenever  an  electric  charge  has  its  rate  of  motion  changed 
the  ether  is  disturbed,  a  certain  amount  of  energy  being  trans- 
ferred or  radiated  from  the  body.  In  the  case  in  which  the 
charged  particle  is  accelerated  for  a  time  dt  with  an  accelera- 

e2a* 
tion  of  "a"  the  energy  radiated  is  E  =  fi  -==-  dt,  where  e  rep- 
resents  the  charge,   and  V   the  velocity  of   light.*   Therefore, 
dividing  by  the  time  during  which  the  change  takes  place,  the 

...  .  ,  .     2e  a 

rate  at  which  energy  is  emitted  is  — —  . 

3V 

Light  production  is  always  associated  with  matter  and  usually 
matter  heated  to  a  high  temperature.  It  is  evident  from  the 
equation  of  the  rate  of  energy  emission  that  since  all  values 
must  be  divided  by  V,  the  velocity  of  light  2.7  X  io10  cm.  per 
second  either  enormously  large  charges  or  very  high  accelerations 
must  be  employed  to  produce  any  appreciable  amount  of  radia- 
tion. Practically,  for  mechanical  reasons,  it  is  impossible  to 
accelerate  charged  bodies  of  a  size  which  can  be  handled,  with 
sufficient  rapidity  to  produce  any  useful  radiation.  Therefore, 
one  must  discover,  if  possible,  some  smaller  units  which  have 
a  higher  ratio  of  charge  to  mass,  than  do  "material"  objects. 

If   the   case   of  bodies   moving  in   simple   harmonic   motion, 

*  J.  J.  Thompson,  Electricity  and  Magnetism. 


402  TRANSACTIONS  I.   E.   S. PART  II 

rotation,  for  instance,  be  considered,  it  will  be  found  that  the 
acceleration  varies  as  the  square  of  the  velocity  of  rotation,  and 

consequently  the  equation  of  energy  radiation  becomes  E  =  — — 

K  per  cycle  where  w  =  the  speed  of  rotation,  and  K  is  constant. 
Also  since  the  acceleration  passes  through  zero  twice  each  revo- 
lution, it  follows  that  the  frequency  of  the  emitted  radiation 
is  w. 

It  is  known  that  for  visible  radiation  w  must  lie  between 
400  X  1012  and  800  X  io12,  the  color  of  the  radiated  light 
depending  upon  the  frequency. 

It  also  follows  that  as  E  =  — ^7-  K  =  the  energy  radiated  per 

3V 

„         teeWKdt  L  .  ,.  .   ,  ,     .- 

cycle,    E*  =  1 ^ —     =    total    energy    radiated,     and     that 

J      3V 

if  there  are  n  charges  being  accelerated  the  total  energy 
from  all  charges  will  be  E/  =  2I — —  or  making  the  constants 

equal  to  C,  E,  =  CXMndt. 

It  is  of  interest  to  note  in  passing,  the  consistency  of  this 
formula  with  that  which  covers  the  radiation  of  energy  from  a 
heated  body  (Stefan's  Law,  which  is  based  on  experiment) 
viz.,  E  =  C  T4  where  T  =  the  absolute  temperature,  in  that  the 
frequency  of  rotation  which  varies  directly  as  the  temperature 
occurs  as  the  fourth  power. 

The  modern  conception  of  matter  assumes  that  the  atom  con- 
sists of  a  central  positive  charge  (which  may  or  may  not  be 
centered  upon  a  material  point)  and  a  number  of  negative 
charges  (or  electrons)  which  are  in  rapid  motion  around  the 
central  charge.  The  charges  are  all  units  and  equal,  the  differ- 
ent properties  of  matter  being  derived  by  the  grouping  and 
numbers  of  the  charges,  and  by  the  groupings  of  the  atoms  and 
molecules. 

Since  then  in  all  matter  there  is  an  unlimited  number  of 
these  groups  of  electrical  charges  moving  about  the  central 
charge  of  opposite  sign,  the  outer  charges  being  in  rapid  motion, 


DARKAH  :     SOME    CONSIDERATIONS    OF    LIGHT    PRODUCTION    403 

one  should  expect,  due  to  the  motion  of  the  charges,  a  continuous 
radiation  of  electromagnetic  waves  from  matter  under  normal 
conditions.  This  is  found  in  all  bodies  and  takes  place  as  the 
so-called  "heat  radiation"  in  accordance  with  Wein's  law,  but  as 
all  surrounding  bodies  are  also  radiating  and  conducting  heat  to 
all  other  bodies,  the  internal  energy  of  the  system  is  not  de- 
pleted. 

If,  however,  the  temperature  of  a  body  be  increased  the  ac- 
celeration of  the  electrons  will  be  increased;  and  thus  also  the 
energy  radiated  and  the  frequency  of  the  radiation  will  be  in- 
creased, until  at  length  those  frequencies  which  lie  in  the 
visible  spectrum  are  reached  and  light  produced.  The  wave- 
lengths of  the  radiations  so  produced  will  cover  the  entire  range 
of  electromagnetic  waves  since  the  apparent  temperature  is  an 
average  of  the  velocities  of  each  atom,  and  the  average  or  ap- 
parent color  will,  of  course,  depend  on  the  average  temperature. 

If  the  temperature  is  raised  until  the  substance  is  vaporized, 
in  other  words,  until  comparatively  unrestrained  motion  can 
take  place  among  the  electrons,  a  condition  resembling  resonance 
is  reached  in  which  the  greater  part  of  the  radiation  of  a  portion 
of  the  atoms  is  confined  to  a  definite  group  of  lines.  This  is 
the  characteristic  of  the  spectrum  observed  when  sodium,  po- 
tassium, calcium,  salts,  etc.,  are  heated  in  a  Bunsen  burner.  The 
effect  of  this  apparent  selective  action  is  to  absorb  energy  of  all 
wave-lengths  and  to  transform  a  considerable  portion  of  this 
absorbed  energy  into  vibrations  of  a  definite  wave-length.  This 
is  what  is  known  as  selective  radiation. 

Returning  to  considerations  of  the  structure  of  the  atom,  it 
has  been  demonstrated*  mathematically  that  assuming  a  given 
number  of  electrons  in  each  atom  moving  under  definite  condi- 
tions they  must  take  definite  configurations.  Since  varying  the 
temperatures  varies  the  speed  with  which  the  electrons  move,  and 
therefore  causes  different  configurations  which  are  due  to  dif- 
ferent rates  of  rotation,  one  should  expect  the  radiant  spectrum 
to  vary  with  the  frequency  of  resonance  for  the  charged  particles. 

An  interesting  converse  effect  of  this  is  the  phenomenon  of 
markedly  increasing  the  motion  of  the  electrons  of  an  atom  by 

*  N.-Bore  and  Albert  Crebore,  Philosophical  Magazine,  July,  1913. 


404  TRANSACTIONS  I.  E.   S. — PART  II 

the  action  of  light  of  a  definite  wave-length.  This  is  applied 
practically  in  photography  where  chemical  reactions  are  caused 
by  the  acceleration  produced  by  light  on  the  electrons  of  the 
silver  atom.  The  photo-electric  effect  where  electrons  are 
actually  torn  from  the  mass  of  a  conductor  in  a  vacuum  by  the 
influence  of  light  of  a  frequency  in  resonance  with  the  motion 
of  the  electrons  is  another  example  of  the  action  of  light  upon 
the  electrons  within  the  atom.  The  effect  of  temperature  to 
raise  the  rate  of  rotation  of  the  electron  sufficiently  to  produce 
resonance  with  the  exciting  wave  is  illustrated  in  those  photo- 
electric cells  which  are  insensitive  to  light  at  ordinary  tempera- 
tures, but  which  become  exceedingly  sensitive  at  an  elevated 
temperature. 

The  theoretically  ideal  illuminant  is  one  which  approximates 
daylight,  and  which  produces  a  spectrum  of  the  frequencies  and 
quantities  necessary  to  approximate  daylight.  It  is  obvious 
therefore,  that  if  it  is  desired  to  confine  the  radiation  to  a  limited 
range,  selective  radiation  must  be  relied  upon.  Since  in  non- 
selective or  temperature  radiation,  due  to  the  constrained  condi- 
tions surrounding  the  motion  of  the  radiators  as  a  result  of  their 
numbers  and  proximity,  the  vibrations  cover  all  frequencies 
and  therefore  the  quantity  of  energy  radiated  in  this  way  at  any 
one  frequency  will  depend  upon  the  number  of  particles  moving 
at  the  necessary  frequency.  Since  there  will  be  relatively  few 
atoms  at  the  extreme  temperatures,  there  will  be  relatively  little 
radiation  at  the  extreme  wave  lengths.  The  greater  part  of  the 
radiation  takes  place  at  the  average  frequencies,  and  a  curve 
plotted  between  wave  length  and  energy  emitted  will  resemble 
the  probability  curve. 

In  choosing  a  measure  of  the  efficiency  with  which  the  energy 
supplied  to  a  body  is  converted  into  useful  visible  radiation,  the 
assumption  is  made  that  all  energy  supplied  is  radiated  in  some 
form.  The  efficiency,  therefore,  becomes  the  ratio  of  the  total 
energy  radiation  to  the  energy  radiated  within  the  range  of 
visibility,  and  is  used  here  in  that  sense. 

A  consideration  of  the  limited  efficiencies  theoretically  possible 
from  incandescence  or  black  body  radiation  alone  will  serve  to 
emphasize  the  advantages  of  selective  radiation.       Thus  assum- 


DARRAH  :     SOME    CONSIDERATIONS    OF    LIGHT    PRODUCTION    405 

ing  a  frequency  of  500  X  JO12  as  an  average  for  normal  visi- 
bility and  taking  Wein's  Law  as  a  basis,  the  energy  at  fre- 

quency  w,   =Ew  =  -^3-  (0      ,     where   the    letters   have    the 

same  significance  as  previously  except  that  C  is  a  constant  equal 
to  about  14,395  and  e  =  the  base  of  the  natural  logarithm  system. 

This  gives  approximately  6,ooo°  absolute  as  the  necessary 
temperature  for  maximum  efficiency  of  radiation  from  an  incan- 
descent body.  The  efficiency  at  this  temperature  while  theo- 
retically about  10  per  cent,  cannot  be  approached  in  actual  prac- 
tice since  the  highest  known  temperature  (the  vaporization  point 
of  carbon)  is  about  3,800°  At  this  temperature  the  efficiency 
of  radiation  of  light  by  incandescence  is  about  5  per  cent,  while 
at  the  melting  point  of  tungsten  the  efficiency  is  about  3  per  cent. 

It  is  evident,  therefore,  that  unless  a  conducting  material  is 
discovered  having  higher  melting  point  than  carbon,  luminous 
efficiencies  from  incandescent  bodies  must  remain  limited.  At 
the  present  time  it  is,  of  course,  the  vaporization  of  the  filament 
of  an  incandescent  lamp  which  limits  the  safe  temperature,  but 
even  when  this  difficulty  has  been  removed  as  seems  commer- 
cially possible  in  some  cases  by  maintaining  a  high  pressure 
around  the  filament,  the  low  melting  point  becomes  a  serious 
limitation  for  which  no  absolute  remedy  is  in  sight. 

The  limitations  of  incandescent  illuminants  may  be  changed, 
of  course,  in  case  a  substance  having  a  radiant  spectrum  lying 
very  largely  within  the  visible  zone  were  discovered — in  other 
words,  one  exhibiting  selective  radiation  in  a  high  degree  while 
solid. 

Referring  again  to  the  equation  previously  derived  for  the 
total    quantity    of    energy    radiated    at    any   given    frequency, — 

E  =     \Ce'7t>*ndt,    and    assuming,    what    is  generally    accepted,* 

that  e  is  the  charge  upon  an  electron,  the  only  variables  are  the 
frequency,  the  time  and  number  of  electrons  involved.  It  fol- 
lows therefore,  from  the  expression  that  the  maximum  intensity 
or  maximum  energy  radiated  by  each  unit  will  be  obtained  with 

*  J.  J.  Thompson,  Conduction  of  Electricity  through  Gases. 
3 


406  TRANSACTIONS  I.  £.   S. — PART  II 

the  highest  frequency  and  when  the  greatest  possible  number  of 
electrons  are  involved. 

In  solid  matter  the  particles  are  too  closely  associated  to  vi- 
brate without  considerable  restraint,  it  is  evident,  therefore,  that 
to  secure  the  effect  of  resonance,  or  selective  radiation,  the  mov- 
ing particles  should  be  as  far  separated  as  possible.  Therefore, 
in  most  cases,  selective  radiation  is  confined  to  gases  or  vapors, 
and  the  study  of  radiation  efficiencies  becomes  the  study  of  gas 
radiation. 

From  previously  quoted  formula,  it  will  be  noted  that  the 
quantity  of  light  radiation  varies  directly  with  the  number  of 
electrons  which  are  subjected  to  the  accelerating  force.  There- 
fore, to  secure  the  maximum  intensity  per  unit  volume  of  radia- 
ting gas,  the  pressure  should  be  as  high  as  possible,  as  this  means 
the  presence  of  a  large  number  of  electron  groups.  The  mere 
presence  of  a  large  quantity  of  electrons,  however,  is  not  in 
itself  an  insurance  of  either  a  high  intensity  of  light  or  a  high 
efficiency.  It  is  only  those  electrons  which  are  subjected  to  a 
change  of  rate  of  motion  (or  acceleration)  between  certain  limits 
which  are  effective  as  radiators.  Since  the  acceleration  of  an 
electron  sufficiently  to  radiate  in  the  wave  lengths  included  in 
the  limits  of  visibility  implies  the  acceleration  of  the  adjacent 
systems  to  nearly  the  same  degree,  a  rough  measure  of  the  effi- 
ciency of  light  production  in  a  gas,  will  be  the  ratio  of  the  num- 
ber of  radiating  systems  to  the  total  number  of  such  systems 
being  acted  upon.  To  secure  a  high  efficiency,  it  is  therefore 
desirable  to  use  as  small  a  quantity  of  gas  as  possible  and  to 
cause  as  many  of  the  electron  systems  as  possible  to  take  part 
in  the  change  in  rate  or  direction  of  motion.  In  passing,  it  is 
of  interest  to  note  that  the  electric  arc  approaches  this  condition 
more  nearly  than  other  illuminants  as  the  pressure  is  compara- 
tively high  and  the  percentage  of  active  particles  relatively  large. 

A  chemical  reaction  between  gases  is  the  best  known  method 
of  accelerating  a  large  percentage  of  the  atoms  present.  The 
percentage  acted  upon  will,  of  course,  be  definitely  determined  by 
the  phase  rule,  and  will  depend  upon  the  rate  of  dissociation  and 
of  recombination  at  the  temperatures  at  which  the  gas  is  worked. 
It  is  of  passing  interest  to  note  that  in  the  various  types  of  flame 


DARRAH  :     SOME    CONSIDERATIONS    OF    EIGHT    PRODUCTION    407 

carbon  arcs  and  metallic  flame  arcs  advantage  is  taken  of  chem- 
ical combination  to  secure  a  high  intensity,  and  a  high  efficiency. 
In  these  arcs  the  various  cerium,  titanium,  or  calcium  compounds 
are  heated  to  the  temperatures  at  which  dissociation  takes  place 
while  the  rapid  movement  of  the  gases  carries  the  dissociated 
products  to  a  region  of  lower  temperatures  where  recombina- 
tion takes  place.  Thus  probably  the  majority  of  the  constituent 
compounds  undergo  a  chemical  reaction  at  least  twice  which  re- 
sults in  a  considerable  acceleration  of  the  electrons,  and  there- 
fore, in  light  production.  The  striking  difference  in  luminosity 
of  different  portions  of  the  flame  arcs,  and  their  sheath-like 
appearance  due  to  the  varying  intensity  of  the  light  in  the  sur- 
rounding layers  is  a  confirmation  of  this  theory,  the  inner  very 
hot  layer  being  the  source  of  relatively  little  radiation  while  the 
surrounding  sheath  which  is  the  seat  of  the  dissociation  and 
recombination  of  the  chemicals  of  the  arc  is  the  source  of  most 
of  the  light. 

It  seems  probable  that  the  high  efficiencies  of  the  Welsbach 
mantle  are  explainable  on  the  same  basis,  namely  that  at  the 
high  temperatures  of  the  gas  flame,  an  unstable  compound  of 
thorium  and  cerium  with  oxygen  or  other  elements  is  formed 
and  that  the  continual  formation  and  decomposition  of  this 
compound  involves  many  more  atoms  (and  thus  effects  many 
more  electrons)  than  would  be  possible  by  merely  heating  the 
mantle.  This  is  in  part  substantiated  by  the  fact  that  heating 
the  mantle  in  an  inert  atmosphere  does  not  produce  very  much 
light.  In  passing  it  is  of  interest  to  note  that  the  addition  of 
thorium  (which  appears  to  be  essential  to  the  successful  opera- 
tion of  the  Welsbach  mantle)  does  not  appreciably  improve  the 
efficiency  or  intensity  of  the  light  from  a  flame  carbon  arc  in 
which  cerium  is  the  chief  illuminant.  A  possible  explanation 
of  this  fact  is  that  the  function  of  the  thorium  is  to  lower  the 
temperature  at  which  the  unstable  compounds  form,  while  in 
the  flame  carbon  arc  the  temperature  is  sufficiently  high  to  make 
this  unnecessary  and  thus  removes  the  advantage  of  thorium. 

In  addition  to  considerations  of  the  intensities  of  the  radia- 
tion, the  position  and  number  of  the  lines  in  the  spectrum  of  a 
substance  is  of  great  importance  in  determining  its  value  as  a 


408  TRANSACTIONS  I.  E.   S. — PART  II 

source  of  light.  Quite  recently  some  convincing  theories  have 
been  developed  which  account  for  the  position  of  the  spectrum 
lines  very  consistently.  These  theories  assume  that  the  atoms 
of  the  various  elements  are  capable  of  existing  in  a  number  of 
more  or  less  stable  states  due  to  the  rate  of  motion  and  the 
number  of  the  component  electrons.  In  passing  from  one  con- 
dition of  stability  to  the  next  stable  condition  a  certain  amount 
of  radiation  is  emitted  which  will  have  a  definite  wave  length 
due  to  the  rate  of  acceleration  of  the  electrons  at  that  instant. 
The  various  stable  states  can  actually  be  demonstrated  by  models 
in  which  small  charged  spheres  are  arranged  around  a  central 
charged   sphere   of   opposite   sign. 

A  mathematical  analysis  of  the  behavior  of  charged  particles 
moving  about  a  central  charged  particle  of  opposite  sign  results  in 
a  formula  from  which  the  calculation  of  the  exact  position  of  the 
spectrum  lines  can  be  made  in  the  case  of  hydrogen  and  some 
of  the  simpler  elements.*  This  theory  leads  to  the  supposition 
that  chemical  affinity  is  the  result  of  electrical  forces  between 
the  groups  of  electrons  and  that  at  the  instant  a  change  from 
one  stable  state  to  another  stable  state  occurs  (or  at  resonance) 
the  chemical  compound  is  destroyed.  On  this  basis  it  would 
seem  that  those  elements  which  have  the  greatest  number  of 
transition  stages  (and  therefore,  have  the  most  intense  and  com- 
plete spectrum)  are  reluctant  to  combine  with  other  elements  and 
would  most  readily  dissociate  when  combined.  This  is  con- 
firmed by  experiment  in  that  hydrogen,  nitrogen,  helium,  argon 
and  neon,  etc.  (the  elements  which  are  perhaps  the  least  active) 
are  peculiarly  efficient  as  illuminants  in  a  Geissler  tube  or  its 
equivalent,  while  oxygen,  flourine,  chlorine  and  the  so-called 
more  active  elements  are  relatively  very  inefficient  as  radiators 
and  have  relatively  very  few  bright  lines  in  their  spectrum. 

SUMMARY. 
Summarizing  the  theories  which  have  been  advanced,  it  ap- 
pears that  we  cannot  hope  to  secure  a  radiation  efficiency  from 
known  incandescent  solids  by  simple  black  body  radiation,  which 
will  be  much  greater  than  5  per  cent.,  due  to  the  low  melting 
points,  and  that  therefore,  the  greatest  expectation  of  obtaining 

*  A.  Crehore,  Philosophical  Magazine,  July,  1913. 


DARRAH  :     SOME    CONSIDERATIONS    OF    EIGHT    PRODUCTION    409 

a  high  efficiency  lies  along  the  lines  of  selective  radiation  from 
electrified  or  incandescent  gases.  Since  the  highest  efficiencies 
will  be  obtained  by  using  the  minimum  amount  of  gas  and  forc- 
ing the  greatest  possible  number  of  electron  systems  to  take  part 
in  the  acceleration,  some  form  of  arc  (or  gaseous  conductor) 
appears  to  be  the  most  promising  field  for  the  future.  Further, 
as  different  materials  have  different  radiant  efficiencies,  the  arc 
should  be  fed  with  materials  which  will  give  the  maximum  radi- 
ation within  the  limits  of  visibility,  and  in  the  proportions  which 
produce  white  light.  In  selecting  the  material  for  supplying 
the  radiation  of  an  arc  it  should  be  borne  in  mind  that  chemical 
inertness  is  frequently  an  indication  of  an  extended  spectrum  of 
considerable  intensity. 

CONCLUSIONS. 
The  conclusions  which  the  facts  summarized  in  this  paper 
seem  to  indicate  are,  that  in  spite  of  the  remarkable  work  which 
has  been  done,  in  raising  the  efficiency  of  the  incandescent  lamp, 
it  is  nearing  its  maximum  theoretical  efficiency.  Although  the 
fact  should  be  kept  in  mind  that  ease  of  application,  size  of  units, 
maintenance  cost  and  first  investment,  may  have  as  much  weight 
in  deciding  upon  the  type  of  illuminant  used  as  the  absolute 
efficiency  and,  therefore,  illuminants  employing  incandescent 
solids  will  undoubtedly  have  a  place  in  the  art  for  a  considerable 
period.  However,  the  vapor  conductor  type  of  illuminant,  as 
for  instance  the  flame  arc  and  metallic  arc,  has  only  begun  to 
approximate  its  possible  efficiencies  and  therefore,  has  a  very 
large  field  ahead  of  it. 


4IO  TRANSACTIONS  I.  E.   S. — PART  II 

THE  PENTANE  LAMP  AS  A  WORKING  STANDARD.* 


BY  E.    C.    CRITTENDEN   AND  A.    H.   TAYLOR. 


Synopsis:  This  paper  recommends  the  use  of  tested  pentane  lamps 
as  secondary  standards  of  candle-power  when  electric  standards  are  not 
available,  and  gives  a  detailed  statement  of  the  method  of  testing  followed 
at  the  Bureau  of  Standards,  with  general  directions  for  the  use  of  the 
lamps.  The  effects  of  variation  in  pentane  and  in  atmospheric  conditions 
are  discussed,  a  correction  for  the  former  is  proposed,  new  determina- 
tions of  the  humidity  correction  factor  are  given,  and  a  chart  is  provided 
to  facilitate  the  reduction  of  observations  to  normal  candle-power. 

INTRODUCTION. 

In  the  Transactions  of  the  Illuminating  Engineering  Society 
there  have  already  been  published  two  papers1  dealing  with  the 
work  on  flame  standards  of  candle-power  which  has  been  done 
at  the  Bureau  of  Standards.  These  papers  were  intended  to  be 
particularly  reports  of  progress  in  the  experimental  work  which 
has  been  done  on  various  types  of  lamps.  The  investigation  of 
the  lamps  was  made  with  the  double  purpose  of  determining 
whether  any  of  them  were  capable  of  furnishing  a  check  on  a 
possible  drift  in  the  value  of  the  unit  now  maintained  by  electric 
incandescent  standards,  and  of  finding  a  satisfactory  working 
standard  for  use  where  electric  standards  are  impracticable. 

The  results  of  that  work  proved  to  be  distinctly  favorable  to 
the  use  of  the  Harcourt  io-candle  pentane  lamp  as  a  secondary 
standard.  Since  the  publication  of  the  papers  mentioned  the  use 
of  such  lamps  has  become  much  more  extensive,  and  the  present 
paper  is  written  with  a  somewhat  different  object  in  view,  namely, 
to  cover  the  questions  which  have  arisen  regarding  their  opera- 
tion. This  involves  a  fulness  of  detail  which  will  not  be  of  much 
interest  to  the  general  reader,  but  which  it  nevertheless  appears 
desirable  to  put  in  print  because  some  of  this  information  is  not 
generally  available. 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 

1  Transactions,  Illuminating  Engineering  Society,  Vol.  V,  pp.  753-778,  1910 ;  and 
Vol.  VI,  pp.  417-432  ;  1911. 


CRITTENDEN   AND  TAYLOR:     PENTANE   LAMP  4II 

Extent  of  Use  of  Pentane  Lamps. — The  number  of  pentane 
lamps  tested  by  the  Bureau  is  not  an  exact  indication  of  the 
extent  to  which  the  use  of  the  lamps  has  increased,  because  many 
have  been  standardized  in  other  laboratories  and  many  others, 
especially  in  the  early  years  of  their  use,  were  not  standardized 
at  all.  Nevertheless  the  number  tested,  as  given  in  the  table 
below,  is  an  indication  of  the  widespread  adoption  of  this 
standard. 

Pentane  Lamps  Tested  by  the  Bureau  of  Standards. 

1908 4 

1909 6 

1910 7 

191 1 17 

I912 35 

1913  (6  months) 15 

Total 84 

Since  in  many  cases  one  of  these  lamps  furnishes  the  basis  on 
which  the  quality  of  the  gas  supply  of  a  city,  so  far  as  candle- 
power  is  concerned,  is  judged,  the  importance  of  the  pentane 
lamp  in  present  practise  will  be  appreciated. 

POSSIBLE  USES  OF  PENTANE  LAMPS. 

While  the  lamp  has  found  its  widest  use  in  gas  testing,  where 
a  flame  standard  is  preferred  because  the  effects  of  atmospheric 
conditions  on  it  compensate  for  similar  effects  on  the  gas  flame, 
its  usefulness  is  by  no  means  limited  to  work  of  that  kind.  The 
corrections  to  determine  the  actual  candle-power  are  easily  made, 
and  the  lamp  affords  a  fairly  convenient  method  of  obtaining  a 
basic  standard  for  laboratories  which  lack  the  batteries  and  accu- 
rate instruments  which  are  necessary  to  get  reliable  values  from 
electric  standards. 

In  some  cases  the  pentane  lamp  may  thus  serve  as  a  basis  of 
standardization  even  when  it  may  be  found  more  convenient  to 
use  other  standards  for  the  actual  tests.  For  example,  when  a 
series  of  measurements  are  made  extending  over  several  hours, 
another  source  such  as  a  kerosene  lamp,  an  Edgerton  standard 
or  a  gas  mantle  may  be  used  as  a  constant  comparison  light  at 
one  end  of  the  bar,  its  value  being  fixed  by  substituting  a  pentane 
lamp  for  the  test  flame  or  lamp  at  the  beginning  and  the  end  of 
the  series.     An  electric  lamp  kept  at  constant  current  may  be 


412  TRANSACTIONS  I.  E-   S. — PART  II 

used  in  the  same  way  as  a  comparison  light.  This  is  exactly  what 
is  done  in  most  portable  photometers,  and  the  calibration  of  such 
photometers  can  be  based  on  a  pentane  lamp  as  well  as  on  an 
electric  standard.  A  modification  of  this  practise  which  is  useful 
where  tests  for  approximate  values  have  to  be  made  at  a  number 
of  stations  is  to  provide  for  each  station  a  portable  standard  unit 
consisting  of  a  low  voltage  lamp  and  a  storage  battery,  the  whole 
unit  being  brought  in  at  regular  intervals  for  recharging  and 
recalibrating  against  the  pentane  standard.  From  a  curve  show- 
ing the  variation  in  candle-power  with  time  of  use  for  such  a 
unit  one  can  obtain  fairly  satisfactory  values  without  the  use  of 
either  regulating  resistances  or  electrical  measuring  instruments. 
The  details  of  such  a  system  in  actual  operation  are  given  by 
Mr.  H.  h.  Farrar,  in  the  Gas  Age  of  April  15,  1913  (pp.  407- 
409).  In  using  such  standards,  as  in  any  case  where  electric  lamps 
are  compared  with  flames,  the  effects  of  humidity  are  important. 
The  readings  can,  however,  be  easily  corrected  for  this  by  the  use 
of  such  a  chart  as  is  given  in  Fig.  3. 

It  is  hardly  necessary  to  give  here  a  long  discussion  of  the 
relative  advantages  of  different  lamps,  and  of  the  reasons  which 
are  leading  to  the  common  use  of  the  pentane  lamp  as  a  working 
standard  in  this  country.  There  are  other  lamps  which  may  be 
used  in  cases  where  their  value  can  be  frequently  checked  or 
where  the  accuracy  required  is  not  great.  In  some  other  cases, 
especially  where  portability  is  an  important  consideration,  the 
Hefner  lamp  is  to  be  recommended;  by  making  many  measure- 
ments under  favorable  conditions  very  good  results  can  be  ob- 
tained with  it,  but  in  general  its  low  candle-power,  comparatively 
red  light,  extreme  sensitiveness  to  drafts,  and  unsteadiness  at 
temperatures  above  70  deg.  F.  make  it  difficult  to  use.  The  pen- 
tane lamp  has  the  disadvantages  of  being  large  and  not  easily 
portable,  of  using  fuel  which  is  expensive  and  somewhat  danger- 
ous, and  of  requiring  more  air  than  ordinary  ventilation  will 
supply.  In  spite  of  these  faults  its  use  is  increasing,  chiefly 
because  it  appears  to  be  the  only  standard  of  candle-power  now 
in  use,  other  than  incandescent  electric  lamps,  which  can  be 
relied  upon  to  give  under  the  usual  working  conditions  the  degree 
of  accuracy  expected  in  present  commercial  practise. 


CRITTENDEN  AND  TAYLOR:     PENTANE  LAMP  413 

STANDARDIZATION  OF  LAMPS  AT  BUREAU  OF 
STANDARDS. 

NECESSITY  OF  PHOTOMETER  TEST. 

The  necessity  for  standardization  of  each  lamp  by  a  photo- 
metric test  arises  from  the  fact  that  different  lamps,  even  of  the 
same  maker  and  supposedly  of  the  same  type,  show  differences 
in  candle-power  which  are  sometimes  as  great  as  4  per  cent. 
The  construction  is  such  as  to  make  it  difficult  to  determine  after 
a  lamp  is  assembled  whether  it  is  built  exactly  according  to 
specifications.  In  fact,  exact  adherence  to  specified  dimensions 
might  not  remove  the  differences  between  lamps  unless  other 
conditions  also  were  definitely  specified.  For  example,  the  cast- 
ings used  in  the  American  lamps  have  a  rough  inner  surface ; 
this  roughness  probably  facilitates  the  transfer  of  heat  to  or  from 
the  air,  and  it  may  also  affect  slightly  the  flow  of  air  through 
the  passages.  Whatever  may  be  the  cause  of  the  differences 
between  lamps,  the  candle-power  of  the  individual  lamp  is  the 
important  thing,  and  hence  the  logical  test  to  make  is  a  deter- 
mination of  the  candle-power.  In  lamps  tested  at  the  Bureau 
details  of  construction  are  not  given  special  attention  unless  some 
part  is  so  made  that  it  appears  to  be  a  possible  cause  of  varia- 
tion in  the  candle-power  of   the  lamp. 

Since  the  lamp  itself  is  only  an  instrument  for  producing  the 
actual  standard,  the  flame,  there  is  little  significance  in  a  value 
given  for  the  lamp  unless  other  factors,  such  as  air  and  fuel 
and  other  conditions  of  operation,  are  known.  The  value  cer- 
tified for  a  lamp  applies  strictly  only  for  its  use  under  conditions 
identical  with  those  under  which  it  was  standardized ;  it  may 
therefore  be  worth  while  to  particularize  rather  fully  the  method 
of  testing  followed  at  the  Bureau. 

PHOTOMETER  ROOM. 
The  photometer  room  used  hitherto  is  approximately  26  x  18  x 
12  feet  (7.92  x  5.49  x  3.66  m.).  Ventilation  is  obtained  by  a 
tempered  air  heating  system.  The  ordinary  ventilation  would  not 
keep  the  air  sufficiently  pure  if  the  flame  gases  were  allowed  to 
escape  into  the  room,  and  consequently  above  the  lamp  is  hung  a 
hood  2  feet  (0.6  m.)  in  diameter  with  an  outlet  into  one  of  the 
ventilating  ducts.    The  draft  into  this  hood  is  just  strong  enough 


414  TRANSACTIONS  I.  E.   S. — PART  II 

to  ensure  that  all  the  gases  rising  from  the  lamp  shall  pass  into 
the  hood  and  thus  be  carried  out  of  the  room.  This  arrange- 
ment is  so  effective  that  no  perceptible  change  in  the  candle- 
power  of  the  lamps  due  to  vitiation  of  the  air  occurs  in  several 
hours  continual  burning.  In  regular  work,  however,  the  room 
is  usually  aired  out  at  least  once  every  hour. 

The  photometer  room  is  in  a  corner  of  the  building  and  often 
has  troublesome  drafts,  in  consequence  of  which  it  has  been  found 
that  the  performance  of  the  lamps  has  been  improved  by  screen- 
ing them  on  three  sides  and  sometimes  on  all  four  sides.  This 
is  done  by  a  wooden  frame-work  about  3  feet  (0.91  m.)  long, 
2  feet  (0.61  m.)  wide  and  3  feet  (0.91  m.)  high,  carrying  a  strip 
of  cloth  which  extends  from  the  top  of  the  frame  down  to  6 
inches  (15.24  cm.)  from  the  bottom,  thus  allowing  free  access  of 
the  air  from  all  sides  below  the  lamp,  but  preventing  any  drafts 
from  striking  the  flame  or  the  circulatory  system  of  the  lamp 
directly.  This  screening  was  adopted  after  many  trials  of  vari- 
ous forms  of  enclosure  had  shown  that  in  most  of  them  the  lamps 
gave  values  slightly  different  from  those  obtained  in  the  open, 
even  when  no  vitiation  of  the  air  in  the  enclosure  could  be 
detected. 

In  special  tests  of  the  state  of  the  air  a  Zeiss  refractometer  has 
been  used  to  determine  the  quantity  of  C02  present.  Measure- 
ments of  atmospheric  moisture  are  regularly  made  with  2 
Assmann  psychrometers,  and  of  pressure  by  a  mercury  barometer. 
The  details  of  the  calculation  of  the  moisture  will  be  given  later 
in  connection  with  the  corresponding  corrections. 

The  photometer  outfit  is  of  the  standard  form  regularly  used 
at  the  Bureau,  with  Lummer-Brodhun  contrast  head  and  a  re-  ' 
cording  device  which  has  been  described  in  the  Bulletin  of  the 
Bureau.2  The  substitution  method  with  constant  intensity  on 
the  disk  is  used  as  in  all  standard  work,  that  is,  a  constant  "com- 
parison lamp"  is  connected  to  the  photometer  head  at  a  fixed 
distance  from  it,  while  the  electric  "working  standards"  and  the 
pentane  lamps  under  test  are  alternately  put  in  on  the  other  side 
for  reading. 

-  G.  W.  Middlekauff,  Bulletin  B.  S.,  7,  p.  iS,  1910.  Reprint  No.  144,  and  Electrical 
World,  Vol.  IvVI,  p.  153,  (July  21)  1910. 


CRITTENDEN  AND  TAYLOR:     PENTANE  LAMP  415 

ADJUSTMENT  OF  LAMPS. 
Before  putting  the  lamp  on  the  photometer  the  chimney  is 
inspected  by  looking  down  through  it  to  see  that  it  is  set  centrally 
over  the  burner,  and  each  time  the  lamp  is  lighted  the  gauge  is 
applied  to  make  sure  that  the  chimney  is  at  the  proper  height, 
that  is,  47  mm.  above  the  burner  when  cold.  The  chimney  seldom 
has  to  be  reset,  but  occasionally  one  creeps  slightly  with  heating 
and  cooling. 

In  setting  up  the  lamp  on  the  table  care  is  taken  to  level  it  so 
that  the  chimney  shall  be  vertical.  The  levels  (plumb  bobs  on 
later  lamps)  attached  for  this  purpose  are  not  always  exact,  and 
this  leveling  is  done  by  dropping  a  plumb  line  from  the  center 
of  the  top  of  the  chimney  so  that  the  bob  should  hang  centrally 
in  the  burner.  The  height  of  the  lamp  is  so  adjusted  that  the 
middle  of  the  flame  is  on  a  level  with  the  center  of  the  photometer 
disk.  The  directions  of  the  Gas  Referees  are  such  as  to  bring 
the  bottom  of  the  chimney  on  a  level  with  the  center  of  the  disk, 
and  this  setting  makes  the  illumination  on  the  disk  follow  the 
inverse  square  law  more  exactly  than  if  the  lamp  is  higher.  The 
American  Gas  Institute's  Committee  on  Taking  Candle-power  of 
Gas  has,  however,  recommended  the  other  position  and  in  order 
to  avoid  confusion  the  Bureau  has  followed  this  recommenda- 
tion. The  lamp  is  usually  placed  so  that  the  line  of  the  bar  passes 
through  the  lamp  standard  as  well  as  through  the  center  of  the 
burner  and  chimney.  In  some  cases,  however,  on  request,  the 
lamps  are  turned  90  deg.  so  that  a  plane  passing  through  the 
chimney  and  the  standard  is  at  right  angles  to  the  line  of  the 
bar.  No  difference  has  been  detected  in  the  values  obtained  in 
the  two  positions  for  any  lamp.  In  either  case  the  chimney  is  so 
turned  that  the  light  from  the  mica  window  does  not  fall  on  the 
photometer  disk  or  on  the  lamp  standard  or  the  pentane  feed 
tube. 

In  measuring  distances  from  the  lamp  the  center  of  the 
burner  is  taken  as  the  point  of  departure.  The  photometric  ob- 
servations are  made  at  a  distance  of  approximately  1  meter.  For 
setting  the  lamp  at  the  proper  point  on  the  scale  it  has  been 
found  convenient  to  follow  the  English  method  of  using  a  rod 
having  near  one  end  a  cylindrical  plug  which  fits  into  the  burner. 


416  TRANSACTIONS  I.  E.   S. — PART  II 

This  rod  is  made  of  such  length  that  the  other  end  just  touches 
the  photometer  head  when  the  center  of  the  burner  is  one  meter 
from  the  disk.  This  fixes  very  exactly  the  point  of  the  scale 
which  is  i  meter  from  the  lamp,  and  a  line  on  the  record  sheet 
is  set  at  this  point.  The  distance  of  the  comparison  lamp  is  then 
so  chosen  as  to  make  the  settings  on  the  pentane  lamp  fall  near 
the  line,  and  in  working  up  the  sheet  the  distances  of  the  groups 
of  points  which  represent  readings  on  the  lamp  are  measured 
from  this  line. 

OPERATION  OF  LAMPS. 

Pentane. — The  pentane  used  is  tested  with  respect  to  density 
and  purity;  the  effects  of  possible  variation  in  the  fuel  are  so 
important  that  a  later  section  of  this  paper  is  devoted  particularly 
to  that  question. 

Flame  Height. — The  method  of  controlling  the  flame  height  has 
been  the  subject  of  considerable  difference  of  opinion.  At  the 
Bureau  no  difference  in  candle-power  has  been  found  to  result 
from  the  use  of  the  various  cocks  at  which  regulation  may  be 
effected.  Even  when  the  saturator  is  entirely  closed  off  from  the 
air,  and  the  pressure  of  the  vapor  in  it  is  allowed  to  run  up  to 
9  inches  (29.53  cm.)  of  water,  regulation  of  the  flame  height 
being  made  by  a  cock  near  the  burner,  no  departure  as  great  as 
1  per  cent,  from  the  normal  candle-power  occurs.  In  standard- 
izing the  older  lamps,  in  which  such  cocks  are  provided,  they  have 
been  used.  The  newer  lamps  do  not  have  these  cocks,  and  flame 
height  is  controlled  by  the  outlet  cock  of  the  saturator,  in  which 
case  greater  ease  of  adjustment  is  obtained  by  clamping  a  lever 
on  to  the  handle  of  the  cock. 

Another  question  concerning  which  there  are  differences  of 
opinion  is  that  of  the  proper  flame  height.  In  English  lamps 
the  top  of  the  flame  should  be  set  at  the  middle  of  the  lower 
mica  window,  about  27  mm.  above  the  bottom  of  the  chimney. 
The  lamps  are  so  proportioned  that  this  height  of  flame  gives  a 
maximum  candle-power  and  a  small  shift  up  or  down  makes  very 
little  change  in  candle-power.  This  is  shown  in  curve  E  of 
Fig.  1  where  abscissas  are  candle-powers  and  ordinates  are 
heights  of  flame  in  the  chimney.  The  effect  of  personal  error  in 
judgment  of  height  of  flame  is  thus  made  very  small. 


CRITTENDEN  AND  TAYLOR:     PENTANE  LAMP 


4*7 


The  American  makers  have  placed  the  crossbar  of  the  window 
lower,  and  direct  that  the  flame  be  set  with  the  tips  at  the  top 
of  the  bar,  which  is  about  22  mm.  above  the  bottom  of  the 
chimney.  As  is  shown  in  curve  A  of  Fig.  1  this  is  not  high 
enough  to  give  the  maximum  candle-power,  and  at  this  height 
small  changes  cause  relatively  large  variation  of  candle-power. 

At  the  Bureau  it  has  been  considered  important  to  retain  the 
advantage  given  by  using  the  maximum  candle-power,  and  inci- 
dentally to  adhere  to  the  original  manner  of  operating  the  lamps. 
The  practise  has  therefore  been  definitely  adopted  of  first  de- 
termining the  height  of  flame  which  gives  the  maximum  candle- 


' 

1           AAMERiCAN  LAf 
EEN6USH  LAMP 

P 

»«»«. 

L. 

mxm 

jj 

H, 

\  \ 

\ 

\  1 

/  i 

r 

___. 

/  / 

"""" 

■OTTO.* 

//     " 

fT? 

* 

' 

\ 

CANDLE-POWER 

Fig.  1.— Variation  of  candle-power  with  pentane  lamp,  with  height  of  flame. 


power  for  a  given  lamp,  and  then  standardizing  the  lamp  with  the 
flame  at  that  height.  With  American  lamps  the  maximum  is 
usually  obtained  when  the  top  of  the  flame  is  about  a  centimeter 
above  the  bar.  A  line  is  marked  on  the  window  to  show  the 
height  used,  and  the  fact  is  stated  in  the  certificate  furnished  with 
the  lamp. 

Time  of  Burning. — The  necessity  of  waiting  a  sufficient  time 
after  lighting  the  lamps  before  making  measurements  was  em- 
phasized in  the  earlier  papers  where  curves  were  given  showing 
the  variation  of  candle-power  with  time  after  lighting  for  English 
and  American  lamps.  It  is  important  to  note  that  the  candle- 
power  rises  rapidly  at  first,  going  above  the  normal,  and  then 
settles  back  to  a  fairly  constant  value.     Considerable  attention 


4l8  TRANSACTIONS  I.  E.   S. — PART  II 

has  been  given  to  the  measurement  of  the  variations  in  tempera- 
ture in  various  parts  of  the  lamp  which  accompany  these  changes 
in  candle-power.  The  details  of  the  measurements  will  not 
be  given  here,  but  it  may  be  said  that  they  have  strengthened  the 
conviction  that  most  of  the  change  in  candle-power  caused  by 
variation  in  conditions  of  operation  may  be  attributed  to  the 
air  circulating  system  and  the  variations  in  the  flow  of  air  which 
arise  from  changes  in  the  relative  temperatures  of  parts  of  the 
lamp.  There  is  a  slow  change  of  temperature  for  a  considerably 
longer  time,  but  the  candle-power  is  practically  constant  after 
30  minutes  in  the  case  of  the  old  style  American  lamps  and  after 
20  minutes  with  the  newer  lamps.  The  excess  of  the  maximum 
over  the  final  candle-power  is  about  3  per  cent,  and  2  per  cent, 
respectively  for  the  two  types.  Lamps  of  the  English  type  run 
from  1  to  iy2  per  cent,  high  at  the  maximum  and  may  be  con- 
sidered constant  after  15  minutes  burning.  The  smaller  heat 
capacity  of  these  lighter  lamps,  which  allows  them  to  reach  a 
steady  state  sooner,  seems  also  to  make  them  more  susceptible 
to  variations  caused  by  changing  drafts. 

BASIS  OF  STANDARDIZATION. 

The  fundamental  unit  in  terms  of  which  lamps  are  certified 
is  the  international  candle3  as  maintained  by  the  primary  elec- 
tric standards  of  the  Bureau.  The  standards  actually  used  in 
tests  are  a  group  of  7  carbon  lamps  operated  at  low  voltage  so 
as  to  match  the  pentane  in  color.  Since  the  flame  is  much  redder 
than  the  ordinary  electric  standards  the  calibration  of  these  spec- 
ial standards  has  taken  considerable  labor  to  insure  that  no 
important  error  be  introduced  because  of  this  color  difference. 
The  group  has  been  compared  with  the  Bureau's  regular  working 
standards  in  three  series  of  measurements  with  a  number  of 
different  observers,  under  the  direction  of  Dr.  Middlekauff,  and 
the  average  values  have  been  as  follows : — 

Candle-power 

I91I 498 

1912 4.98 

1913 4-97 

It  appears  therefore  that  the  uncertainty  in  the  value  of  these 

3  Bureau  of  Standards,  Circular  No.  15. 


CRITTENDEN  AND  TAYLOR:     PENTANE  LAMP  419 

standards  cannot  be  very  great.  As  a  check  on  their  permanence 
another  group  carefully  compared  with  them  is  preserved  and 
used  only  for  checking  the  working  standards.  A  similar  group 
has  also  been  prepared  to  be  sent  to  England  in  order  to  obtain 
a  direct  comparison  between  the  standards  of  the  National 
Physical  Laboratory  and  those  of  the  Bureau  at  pentane  color. 

To  obtain  the  normal  candle-power  of  the  flame,  corrections 
are  made  for  the  effects  of  atmospheric  moisture  and  pressure. 
These  corrections  will  be  considered  in  a  separate  section  since 
they  must  be  used  whenever  it  is  desired  to  obtain  absolute  candle- 
power  values  with  a  flame  standard.  As  a  check  against  the 
introduction  of  errors  by  other  conditions,  such  as  poor  ventila- 
tion, which  might  affect  the  flames  but  not  the  electric  standards, 
a  pentane  lamp  of  known  candle-power  is  always  included 
in  tests. 

NUMBER  OF  MEASUREMENTS. 

The  number  of  measurements  made  on  each  test  lamp  in 
the  course  of  standardization  depends  somewhat  on  the  con- 
sistency of  the  results  obtained.  Usually  each  lamp  is  placed  on 
the  photometer  about  six  times,  and  each  time  four  groups  of 
sets  are  made,  each  group  consisting  of  25  to  40  settings  of  the 
photometer.  The  candle-power  result  for  each  group  is  worked 
up  separately,  and  if  the  average  deviation  of  these  values  from 
the  final  mean  is  not  materially  greater  than  one-half  per  cent., 
and  the  values  obtained  for  the  check  lamp  run  in  the  test  are 
also  normal,  the  test  is  considered  satisfactory.  If  these  condi- 
tions are  not  fulfilled,  further  measurements  are  made  until  it 
is  believed  that  the  mean  candle-power  of  the  lamp  is  sufficiently 
well  determined.  The  candle-power  is  certified  to  the  nearest 
tenth  of  a  candle  unless  the  average  of  test  results  falls  very 
nearly  half  way  between  tenths,  in  which  case  a  subscript  5  is 
given  in  the  hundredths  place;  this  is  written  as  a  subscript  to 
indicate  that  it  is  not  considered  as  definitely  established,  but 
merely  as   representing  the  average  of  test  results. 

ACCURACY  OF  VALUES. 

With  all  the  precautions  taken  it  is  believed  that  the  values 
certified  for  the  lamp  are  correct  within  one  per  cent.,  that  is, 


420  TRANSACTIONS  I.  E.   S. — PART  II 

that  under  similar  conditions  the  lamps  should  give  average 
results  within  one  per  cent,  of  those  certified.  It  has  occasionally 
happened  that  a  second  test  on  a  lamp  has  given  a  result  fully- 
one  per  cent,  different  from  the  value  originally  certified.  The 
result  obtained  from  a  single  time  on  the  photometer  sometimes 
departs  as  much  as  2  per  cent,  from  the  mean,  but  the  maximum 
deviation  is  seldom  as  great  as  this.  As  for  the  permanence  of 
the  calibration  no  evidence  has  been  obtained  indicating  any 
appreciable  change  with  time.  For  the  Bureau's  two  Chance 
lamps  which  are  regularly  used  as  checks  on  tests  the  average 
candle-powers  found  in  1910  were  9.87  and  9.89;  the  averages 
for  the  first  six  months  of  1913  have  been  9.90  and  9.87. 

Conditions  in  different  laboratories  are  not  likely  to  be  exactly 
the  same,  and  in  order  to  depend  on  reproducing  values  in  dif- 
ferent places  as  closely  as  one  per  cent,  one  would  probably  need 
to  give  some  care  to  reproducing  conditions  of  operation.  It 
would  seem,  however,  that  under  any  reasonably  good  conditions 
the  difference  ought  to  be  well  within  2  per  cent.,  for,  using  good 
pentane  and  correcting  for  atmospheric  moisture  and  pres- 
sure, it  is  difficult  to  produce  that  much  variation  in  candle- 
power  by  any  intentional  change  of  conditions  except  vitiation  of 
the  air  or  incorrect  flame  height. 

GENERAL  DIRECTIONS  FOR  USE  OF  PENTANE 

LAMPS. 

It  is  hardly  practicable  to  give  here  detailed  directions  as  to 
the  exact  procedure  to  be  followed  in  any  case,  but  attention  may 
be  called  in  a  general  way  to  some  precautions  which  should  be, 
taken  in  using  the  lamps.  Many  of  the  details  of  adjustment 
and  operation  which  have  already  been  discussed  will  be  merely 
mentioned  here.  The  best  results  will  presumably  be  obtained 
by  following  the  same  methods  in  operating  a  lamp  as  were  used 
in  standardizing  it.  If  the  lamp  to  be  used  has  been  standardized 
at  the  Bureau  a  reference  to  the  preceding  pages  will  answer 
most  questions  as  to  proper  procedure.  If  the  value  of  the  lamp 
has  been  assigned  by  another  laboratory  such  questions  should 
usually  be  referred  to  that  laboratory. 


CRITTENDEN   AND  TAYLOR:     PENTANE  LAMP  421 

VENTILATION  AND  EFFECTS  OF  VITIATION  OF  AIR. 
Since  all  flames  depend  upon  combustion  it  is  to  be  expected 
that  their  intensity  will  vary  with  the  proportion  of  oxygen  and 
of  other  constituents  in  the  air  supplied  to  the  gaseous  fuel  of  the 
flame.  To  maintain  a  flame  in  a  constant  condition  requires, 
therefore,  not  only  uniform  fuel,  but  also  a  uniform  proportion 
of  oxygen  in  the  air  supplied  to  the  flame.  Good  ventilation  is 
consequently  desirable  for  all  work  with  flames,  and  is  indis- 
pensible  where  lamps  of  other  kinds  are  to  be  compared  with  a 
flame  standard.  The  quantity  of  air  which  must  be  supplied  de- 
pends largely  on  the  size  and  number  of  flames  in  the  room  and 
also  on  the  method  of  ventilating.  The  flames  themselves  set  up 
currents  of  air,  and  if  possible  these  currents  should  be  utilized 
to  remove  immediately  from  the  room  all  the  vitiated  air  coming 
from  the  lamps.  A  hood  placed  above  the  photometer  with  a 
rising  pipe  to  lead  the  warm  air  out  of  the  room  is  most  effective. 

If  the  gases  from  the  flames  are  allowed  to  diffuse  into  the  room 
and  are  simply  diluted  by  the  air  entering,  a  much  larger  amount 
of  fresh  air  will  be  needed.  In  such  a  case  it  has  been  esti- 
mated4 that  to  keep  the  carbon  dioxid  content  of  the  air  down  to 
six  parts  in  10,000,  3,000  cubic  feet  (84.95  cu.  m.)  of  fresh 
air  per  hour  must  be  supplied  for  each  person  in  a  room. 
A  committee  of  the  American  Gas  Institute5  has  on  this 
basis  estimated  that  36,900  cubic  feet  (1045  cu.  m.)  of  air  per 
hour  should  be  supplied  for  a  photometer  room  where  two 
persons  work  with  a  pentane  lamp  and  a  5-foot  (141.6  1.)  gas 
flame,  since  in  the  production  of  carbon  dioxid  the  gas  flame  is 
approximately  equivalent  to  6  persons  and  the  pentane  lamp  to 
four. 

In  a  room  where  air  enters  at  the  bottom  and  escapes  at  or 
near  the  top,  the  quantity  estimated  as  above  (about  12,000  cubic 
feet  (340  cu.  m.)  per  hour  for  a  pentane  lamp)  is  probably 
ample  for  any  purpose,  and  if  care  is  taken  to  allow  direct  escape 
of  the  flame  gases  a  much  smaller  amount  is  sufficient. 

Some  caution  is  necessary  in  making  calculations  based  on  the 
quantity  of  carbon  dioxid  in  the  air.     Experimental  determina- 

4  Kent,  Mech.  Eng.  Pocketbook  ;  1912  Ed.,  p.  654. 
6  Proceeding!  A  met  ican  Gas  Institute,  Vol.  II,  pp.  481-4S2,  1907. 
4 


422  TRANSACTIONS  I.   E.   S. — PART  II 

tions  of  the  effect  on  flames  have  sometimes  been  made  by  adding 
carbon  dioxid  to  the  air,  and  it  must  be  remembered  that  the 
dilution  of  the  air  by  a  quantity  of  added  C02  has  relatively  little 
effect  as  compared  with  the  conditions  arising  when  the  same 
amount  is  formed  in  the  process  of  combustion,  using  up  the 
oxygen  of  the  air.  In  the  formation  of  five  cubic  feet  of  C02 
by  the  combustion  of  pentane,  for  example,  8  cubic  feet  of 
oxygen  are  used.  Since  the  abstraction  of  i  cubic  foot  of 
oxygen  has  practically  the  same  effect  on  the  composition  of  the 
air  as  the  addition  of  5  cubic  feet  of  inert  gas,  it  is  evident  that 
the  reduction  in  the  amount  of  the  active  oxygen  is  of  much 
greater  importance  than  the  increase  in  the  diluting  gases.  The 
effect  produced  on  the  Hefner  lamp  by  adding  one  part  of  C02 
to  1,000  of  air,  according  to  Liebenthal,6  is  a  reduction  of  0.7  per 
cent,  in  the  candle-power,  but  the  generation  of  the  same  propor- 
tion of  C02  by  combustion  and  breathing  in  a  room  has  been 
found7  to  be  accompanied  by  a  decrease  of  2.2  to  3  per  cent,  in 
the  Hefner,  the  pentane  lamp  and  other  flames.  In  general  the 
effect  depends  on  the  manner  of  production  of  the  C02,  since 
the  greater  part  of  the  decrease  is  due  not  to  the  presence  of  C02 
but  to  a  deficiency  of  oxygen.  Determinations  of  the  amount  of 
C02  in  the  air  consequently  do  not  furnish  sufficient  data  for  exact 
correction  for  vitiation  of  the  air,  but  are  useful  in  that  they 
enable  one  to  judge  the  effectiveness  of  the  ventilation. 

In  many  cases,  however,  determinations  of  C02  are  not  prac- 
ticable and  the  effectiveness  of  the  ventilation  must  be  judged 
by  other  means.  When  facilities  are  available  for  setting  an 
electric  lamp  repeatedly  to  the  same  current  a  direct  test  can  be 
obtained  by  making  a  series  of  measurements  of  the  electric  lamp 
against  the  pentane  and  finding  whether  the  latter  shows  a  grad- 
ual decrease  after  it  should  have  reached  a  constant  value. 
Another  test  which  is  as  definite  as  determination  of  the  carbon 
dioxid  and  far  more  easily  carried  out  is  afforded  by  careful 
measurement  of  the  humidity,   for  the  processes  which  use  up 

*  Zeitsch.f.  Ivstrumentenkunde,  Vol.  XV,  p.  157,  1895. 

"  C  C.  Paterson,  Collected  Researches,  National  Phys.  Lab.,  Vol.  Ill,  p.  49,  1908. 
Butterfield,  Haldane  &  Trotter,  Journal  Gas  Lighting,  Vol.  CXV,  p.  290,  1911,  and 
American  Gas  Light  Journal,  Vol.  XCV,  p.  145,  1911.     Also  unpublished  tests  of  Bureau  of 
Standards. 


CRITTENDEN  AND  TAYLOR:     PENTANE  LAMP  423 

oxygen  add  water  as  well  as  C02  to  the  air.  If  the  water  vapor 
regularly  increases  by  an  appreciable  amount  during  the  opera- 
tion of  the  lamp  the  ventilation  is  not  satisfactory.  An  increase 
of  one  liter  of  water  vapor  per  cubic  meter  of  air,  when  caused 
by  poor  ventilation,  is  quite  regularly  accompanied  by  such  vitia- 
tion as  to  cause  about  2  per  cent,  decrease  in  the  candle-power 
of  flames,  in  addition  to  the  0.6  per  cent,  decrease  caused  by  the 
water  vapor  itself. 

In  designing  ventilating  inlets  ample  capacity  should  be  pro- 
vided to  allow  a  slow  flow  of  air  into  the  room  in  order  to  avoid 
drafts  which  would  cause  unsteadiness  of  the  flame.  Similarly 
if  a  hood  is  used  as  recommended  the  outlet  must  be  so  arranged 
that  the  flow  of  air  into  the  hood  shall  not  be  too  vigorous. 
Troublesome  drafts  are  apt  to  arise  if  the  walls  of  the  photometer 
room  differ  much  in  temperature  from  the  air.  It  is  therefore 
desirable  that  none  of  them  shall  be  exterior  walls  of  the  building. 
If  such  walls  are  unavoidable  they  should  be  either  jacketed 
with  non-conducting  material  or  covered  by  a  false  wall  with 
an  air-space.  If  possible  the  room  should  be  so  free  from  drafts 
that  the  flame  will  burn  steadily  without  other  protection  than 
the  necessary  photometric  screens.  If  further  protection  from 
drafts  is  necessary  a  screen  of  the  form  described  on  a  preceding 
page  should  be  used. 

PREPARATION  OF  LAMP. 

In  preparing  the  lamp  for  use  one  should  observe  the  following 
details :  centering  of  chimney  over  the  burner,  height  of  chimney 
above  the  burner,  direction  in  which  the  mica  window  is  turned, 
amount  of  pentane  in  the  saturator,  height,  orientation  and  level- 
ing of  the  lamp,  and  its  distance  from  the  middle  point  of  the  bar 
or  some  other  definite  point  on  the  scale.  It  is  advisable,  at  least 
in  the  beginning,  to  check  the  last  two  adjustments  by  direct 
measurement  previously  described.  If  then  the  plumb  bobs  (or 
the  level  and  bob)  are  found  to  be  correct  they  may  be  used 
thereafter. 

The  saturator  should  be  from  one-third  to  two-thirds  full  of 
pentane  at  starting,  and  the  height  of  the  liquid  as  seen  against 
the  window  of  the  saturator  should  never  be  less  than  */£  inch 
(3.175  mm.).     Since  pentane  is  very  volatile  and  inflammable, 


424  TRANSACTIONS  I.  E.   S. — PART  II 

and  the  heavy  vapor  flows  downward,  it  is  extremely  hazardous 
to  fill  a  lamp  while  it  is  burning.  This  should  never  be  attempted 
under  any  circumstances,  and  it  should  be  an  inviolable  rule  of 
the  laboratory  that  no  pentane  in  an  open  vessel  be  brought  near 
any  flame. 

OPERATION. 

Immediately  before  the  lamp  is  lighted  the  height  of  the  chim- 
ney should  always  be  tested  with  the  gauge,  and  occasionally  the 
gauge  itself  should  be  measured  to  see  that  it  remains  47  mm. 
long.  To  light  the  lamp  open  first  the  saturator  inlet  cock;  then 
holding  a  lighted  match  over  the  burner  open  the  outlet  cock 
gradually.  If  the  lamp  has  a  regulating  cock  near  the  burner, 
the  saturator  cocks  may  both  be  opened  and  the  flame  controlled 
by  this  regulating  cock.  If  no  such  cock  is  provided  the  saturator 
outlet  cock  is  used  for  regulation.  Usually  on  opening  the  cocks 
the  vapor  will  flow  so  that  the  lamp  will  light,  but  it  is  sometimes 
necessary  to  start  the  flow  of  vapor  by  blowing  gently  into  the 
saturator  inlet.  Sometimes  also  the  lamp  at  first  burns  with  a 
small  blue  flame  because  the  heavy  vapor  has  flowed  down  into 
the  burner  and  prevents  the  normal  circulation  of  air  from  start- 
ing. If  this  happens  shut  off  the  cocks,  and  blow  up  into  the 
outer  chimney  to  start  the  air,  keeping  below  the  burner  to  avoid 
the  small  burst  of  flame  which  may  result  if  the  flame  has  not 
entirely  died  out  in  the  tip.  The  lamp  will  then  light  normally. 
After  it  is  lighted  see  that  the  conical  hood  around  the  flame  is 
so  placed  that  the  whole  flame  is  visible  from  the  photometer 
disk. 

The  flame  should  be  kept  at  approximately  the  correct  height, 
and  no  measurements  should  be  made  till  the  lamp  has  burned 
15,  20,  or  30  minutes  according  to  whether  it  is  of  the  English 
type  or  of  the  new  or  old  American  forms.  Unless  otherwise 
specifically  stated  in  the  lamp  certificate  for  the  American  lamps 
the  proper  flame  height  is  obtained  when  the  tips  are  just  above 
the  crossbar  of  the  window,  while  in  English  lamps  the  flame 
should  extend  half-way  up  the  lower  window.  As  already  stated 
American  lamps  standardized  recently  at  the  Bureau  of  Standards 
have  a  line  on  the  window  to  which  the  top  of  the  flame  is  to  be 
set.    After  the  lamp  has  burned  a  few  minutes  such  changes  in 


CRITTENDEN  AND  TAYEOR :     PENTANE  LAMP  425 

flame  height  as  occur  are  gradual,  and  except  in  the  most  careful 
work  it  is  not  necessary  to  have  a  special  observer  to  watch  the 
flame. 

The  lamp  is  extinguished  by  shutting  off  the  saturator  cocks. 
When  it  is  not  in  use  both  cocks  should  be  kept  closed  and  a  cap 
should  be  placed  over  the  burner  to  prevent  injury  to  it  or  the 
collection  of  dust  in  the  passages.  All  parts  of  the  lamp  should 
be  kept  well  blackened. 

PENTANE:  PREPARATION,  TESTING  AND  USE. 

The  directions  of  the  Gas  Referees  for  the  preparation  and 
testing  of  pentane  are  as  follows : 

Preparation. — Light  American  petroleum,  such  as  is  known  as  Gasoline 
and  used  for  making  air-gas,  is  to  be  further  rectified  by  three  distilla- 
tions, at  55  deg.  C,  50  deg.,  and  45  deg.  in  succession.  The  distillate  at 
45  deg.  is  to  be  shaken  up  from  time  to  time  during  two  periods  of  not 
less  than  3  hours  each  with  one-tenth  its  bulk  of  (1)  strong  sulphuric 
acid,  (2)  solution  of  caustic  soda.  After  these  treatments  it  is  to  be 
again  distilled,  and  that  portion  is  to  be  collected  for  use  which  comes 
over  between  the  temperatures  of  25  deg.  and  40  deg.  It  will  consist 
chiefly  of  pentane,  together  with  small  quantities  of  lower  and  higher 
homologues  whose  presence  does  not  affect  the  light  of  the  lamp. 

Testing. — The  density  of  the  liquid  pentane  at  15  deg.  C.  should  not 
be  less  than  0.6235  nor  more  than  0.626  as  compared  with  that  of  water 
of  maximum  density.  The  density  of  the  pentane  when  gaseous,  as  com- 
pared with  that  of  hydrogen  at  the  same  temperature  and  under  the  same 
pressure,  may  be  taken.  This  is  done  most  readily  and  exactly  by  Gay 
Lussac's  method,  under  a  pressure  of  about  half  an  atmosphere  and  at 
temperatures  between  25  deg.  and  35  deg.  The  density  of  gaseous  pen- 
tane should  lie  between  36  and  38. 

Any  admixture  with  pentane  of  hydrocarbons  belonging  to  other  groups 
and  having  a  higher  photogenic  value,  such  as  benzene  or  amylene  must 
be  avoided.  Their  presence  may  be  detected  by  the  following  test. 
Bring  into  a  stoppered  4-oz.  bottle  of  white  glass  10  cc.  of  nitric  acid, 
specific  gravity  1.32  (made  by  diluting  pure  nitric  acid  with  half  its  bulk 
of  water)  ;  add  1  cc.  of  a  dilute  solution  of  potassium  permanganate, 
containing  0.1  gram  of  permanganate  in  200  cc.  Pour  into  the  bottle 
50  cc.  of  the  sample  of  pentane,  and  shake  strongly  during  five  successive 
periods  of  20  seconds.  If  no  hydrocarbons  other  than  paraffins  are 
present,  the  pink  color,  though  somewhat  paler,  will  still  be  distinct;  if 
there  is  an  admixture  of  as  much  as  1/2  per  cent,  of  amylene  or  benzene, 
the  color  will  have  disappeared. 

For  the  benefit  of  those  who  are  not  chemists  it  should  be  said 
that  this  last  test  should  always  be  preceded  by  a  blank  test  to 


426  TRANSACTIONS  I.  E.   S. — PART  II 

check  the  purity  of  the  reagents,  for  if  the  nitric  acid  is  not  pure 
it  will  decolorize  the  permanganate.  The  acid  should  be  kept  in 
the  dark  to  avoid  deterioration. 

The  pentane  used  at  the  Bureau  is  tested  with  respect  to  purity 
and  density  of  the  liquid.  The  density  is  conveniently  deter- 
mind  by  a  hydrometer  carrying  a  thermometer.  The  correction 
for  temperature  is  important,  being  about  0.00 1  per  degree  Centi- 
grade or  0.00055  Per  degree  F.  In  the  sealed  ether  cans 
in  which  pentane  is  now  purchased  no  difficulty  has  been  found 
in  keeping  a  supply  which  fulfills  the  specifications  when  the  cans 
are  opened,  but  the  more  rapid  evaporation  of  the  lighter  frac- 
tions raises  the  density  as  the  fuel  is  consumed  in  the  lamp.  It 
is  quite  impracticable  to  work  with  pentane  within  the  limits  pre- 
scribed, for  the  density  is  certain  to  be  too  high  before  one-tenth 
of  the  pentane  is  consumed;  in  fact  the  density  usually  reaches 
0.635  when  a  little  over  half  has  been  used. 

CORRECTION  DETERMINED  BY  DENSITY. 

The  directions  of  the  London  Gas  Referees  are  to  empty  the 
saturator  completely  at  least  once  a  month  when  three  tests  daily 
are  made,  but  this  appears  to  be  decidedly  too  long  a  period.  The 
residue  of  higher  density  gives  a  slightly  higher  candle-power 
than  the  fresh  pentane.  When  repeated  additions  of  fresh  fuel 
are  made,  and  the  accumulated  residue  remains  in  the  lamp,  a 
mixture  is  obtained  such  that  an  appreciable  change  in  candle- 
power  may  occur  in  a  relatively  short  time.  It  therefore  appears 
desirable  to  empty  the  saturator  after  it  has  been  replenished  only 
three  or  four  times.  It  is  not  necessary,  however,  to  discard  the 
portions  emptied  out ;  if  these  are  collected  and  not  mixed  with 
fresh  pentane,  fairly  reliable  results  can  be  obtained  by  using  the 
residues  thus  collected,  making  a  suitable  correction  determined 
by  the  density.  Considerable  attention  has  been  given  to  the 
determination  of  such  a  correction,  and  the  results  of  the  meas- 
urements made  are  shown  in  Fig.  2.  The  relation  between  candle- 
power  and  density  is  probably  not  really  linear,  but  for  present 
purposes  may  be  assumed  so.  It  will  be  seen  that  the  change  in 
candle-power  averages  about  1  per  cent,  for  an  increase  of  0.01 
in  the  density  of  the  fuel.  Samples  of  pentane  from  different 
sources  which  have  been  initially  high  in  density  have  also  given 


CRITTENDEN  AND  TAYLOR:     PENTANE  LAMP 


427 


results  agreeing  fairly  well  with  this  rule.  Consequently  it 
appears  to  be  allowable  to  apply  such  corrections  in  many  cases 
where  the  expense  for  pentane  can  thus  be  cut  down  materially 
without  any  considerable  sacrifice  of  accuracy. 

In  the  standardization  of  lamps  at  the  Bureau  the  pentane  used 
is  always  kept  below  0.635  in  density.  The  values  certified  may 
therefore  be  considered  as  correct  for  a  density  of  0.630.  No 
correction  need  be  made  for  pentane  below  0.635,  but  above  that 
density  correction  is  desirable.  For  example,  if  the  pentane  used 
runs  up  to  0.650  the  candle-power  obtained  is  presumably  2  per 
cent,  above  that  certified  for  the  lamp,  and  this  density  is  likely 


010.0 


3.6 


• 

0 

v^ 

• 

^ 

$x 

-  ; 

°°° 

.625  .635  .645  .655  .665 

DENSITY  OF  PENTANE    AT  59°  FAHR. 

Fig.  2. — Effect  of  pentane  density  on  candle-power  of  pentane  lamps. 


to  be  reached  if  a  gallon  of  pentane  is  used  without  removing 
the  residue  from  the  saturator. 

It  should  be  noted  that  the  above  discussion  refers  to  pentane 
which  is  initially  fairly  pure,  as  shown  by  the  permanganate  test. 
If  the  fresh  pentane  contains  impurities  these  are  usually  con- 
centrated by  the  fractional  distillation  in  the  lamp,  and  the  result 
may  be  an  increase  in  candle-power  greater  than  that  given  above. 

EFFECTS  OF  ATMOSPHERIC  CONDITIONS. 
It  is  usually  assumed  that  the  effect  of  all  atmospheric  condi- 
tions is  the  same  on  all  flames,  and  is  therefore  automatically 
corrected  when  flames  are  compared  with  flames,  but  it  is  some- 
times  necessary  to   know   the   amount   of   departure   from   the 


428  TRANSACTIONS  I.  E.   S. — PART  II 

normal  value,  as  when  flame  standards  are  compared  with  electric 
lamps.  The  dependence  of  flame  intensity  on  the  composition  of 
the  air  is  so  complicated  a  matter  that  the  determination  of  exact 
corrections  for  vitiation  of  the  air  is  impracticable,  and  when 
observations  are  to  be  reduced  to  normal  values  good  ventilation 
is  indispensable  so  that  no  such  corrections  shall  be  needed. 
With  pure  air,  however,  there  are  considerable  variations  in 
candle-power  because  of  variations  in  the  amount  of  moisture  and 
in  the  barometric  pressure. 

TEMPERATURE. 
Temperature  might  also  be  expected  to  affect  the  pentane  lamp 
appreciably,  but  nearly  all  investigators  have  agreed  that  within 
the  usual  laboratory  range  it  does  not  do  so.  In  the  work  at  the 
Bureau  there  has  so  far  been  available  no  means  of  changing 
temperature  and  humidity  independently,  and  the  effects  of  the 
two  cannot  be  separated  with  certainty.  The  results  obtained  can 
be  represented  about  equally  well  by  assuming  that  temperature 
has  no  effect  or  by  making  a  small  correction  for  temperature 
and  using  a  correspondingly  different  factor  for  water  vapor.  In 
accordance  with  the  usual  custom  the  former  practise  will  be  fol- 
lowed, and  the  present  discussion  will  be  limited  to  the  effects  of 
barometric  pressure  and  of  water  vapor. 

BAROMETRIC  PRESSURE. 
In  general,  flames  give  less  light  when  the  barometric  pressure 
is  low,  but  the  effects  on  different  sorts  of  lamps  are  markedly 
different.  For  the  10  candle-power  pentane  lamp  a  change  of 
0.6  per  cent,  per  cm.  (1.5  per  cent,  per  inch)  was  found  at 
the  German  Reichsanstalt,s  while  the  result  obtained  at  the 
English  National  Physical  Laboratory9  is  0.8  per  cent,  per  cm. 
(2  per  cent,  per  inch),  and  more  recent  work10  in  England 
has  verified  the  latter  value  for  pressures  near  the  normal.  Vari- 
ous determinations  at  the  Bureau  have  not  been  consistent,  chiefly 
because  of  the  small  range  of  pressure  obtained.  The  results  for 
the  determinations  which  should  have  been  most  reliable  have 

8  Jour,  fur  Gas.  u.  Wasser.,  Vol.  XL,IX,  p.  561.  1906. 

9  Electrician  (London),  Vol.  UJI,  p.  571,  1904.  Journal  Institution  of  Electrical  Engineers , 
Vol.  XXXVIII,  p.  271,  1906-7.  Journal  Gas  Lighting,  Vol.  XCIX,  p.  232,  1907.  N.  P.  t,. 
Collected  Researches,  Vol.  Ill,  p.  49,  1908. 

10  Butterfield,  Haldane,  and  Trotter,  Journal  Gas  Lighting,  Vol.  CXV,  p.  290,  1911,  and 
Ameiican  Gas  Light  Journal,  Vol.  XCV,  p.  145,  1911. 


CRITTENDEN"   AND  TAYLOR:     PENTANE  LAMP 


429 


varied  from  0.6  to  0.8  per  cent.,  and  consequently  the  English 
value,  0.8  per  cent,  per  cm.  has  been  used.11  The  chart  of  devia- 
tions given  in  Fig.  3  is  plotted  on  this  basis. 

WATER  VAPOR  DETERMINATIONS. 

Water  vapor  in  the  air  lowers  the  intensity  of  the  flame;  the 

effect  has  been  found  to  be  proportional  to  the  amount  of  water 

present.     This  amount  is  expressed  in  liters  (/)  of  water  vapor 

per  cubic  meter  of  dry  air,  in  other  words  in  parts  of  water 


4 

[DRY  BULB  TtMftRATURE6' 

TT 

1  i 

: 

w 

fj 

// 

! 

!« 

/> 

j 

z 
< 

4 

< 

!< 

z 
g« 

g" 

[ 

7 

If 

u 

i 

i\ 

R! 

- 

:■: 

1 

/ 

7 

h 

/ 

L 

n 

1 1 

H 

1  f 

1 

i 

/ 

1 

j  J 

7 

b   s 

AODIT 
DUE 

S     5    2     1  -0< 

IOBHWETER 

/ 

1 

/ 

f-'ig.  3- 


Variation  of  candle-power  of  pentane  lamps  with  humidity 
and  barometric  pressure. 


vapor  per  1,000  parts  of  dry  air.  The  pressure  (e)  of  the  water 
vapor  is  determined  by  means  of  a  hygrometer  (preferably  of 
the  ventilated  type)  using  tables  adapted  to  the  particular  type 
of  hygrometer.    If  the  barometric  pressure  is  represented  by  b, 

e 


1  = 


X  1,000. 


b  —  e 

A  normal  value  (n)  for  the  amount  of  water  vapor  must  be 
chosen  more  or  less  arbitrarily.  Then  if  I„  is  the  normal  inten- 
sity of  a  given  flame  (that  is,  the  intensity  when  there  are  n  liters 
of  water  vapor  per  cubic  meter  of  dry  air),  the  intensity  I  at  any 
particular  time  is  given  by  the  equation. 

1  =  1,  [1+  (;;  —  /)  a]. 

11  In  the  Transactions  of  the  Illuminating  Engineering  Society.  Vol.  V,  p.  776,  1910, 
it  was  inadvertently  stated  that  the  factor  0.6  is  used,  when  it  should  have  been  said  that 
this  factor  was  obtained  from  the  data  then  reported  on. 


430  TRANSACTIONS  I.  E.   S. — PART  II 

The  normal  proportion  of  water  vapor  has  been  fixed  at  8.0  liters 
for  the  pentane  lamp ;  the  correction  factor  a  has  been  several 
times  determined,  but  the  agreement  between  results  in  different 
laboratories  is  not  close.  These  determinations  were  rather  fully 
discussed  in  one  of  the  earlier  articles  mentioned12 ;  it  is  sufficient 
to  recall  here  that  the  National  Physical  Laboratory13  found  a  to 
be  0.0066,  whereas  the  Reichsanstalt13  obtained  0.0055  and  the 
Bureau  of  Standards  0.0057.  The  more  recent  English  tests13 
have  given  0.00625. 

The  method  of  testing  lamps  at  the  Bureau  furnishes  continual 
data  for  the  redetermination  of  the  correction  factor,  and  com- 
plete calculations  have  been  made  using  the  results  up  to  the 
beginning  of  the  present  year  on  all  lamps  whose  tests  have 
included  a  range  of  5  liters  of  water  vapor  or  more.  To  reduce 
the  labor  of  calculation  the  values  obtained  each  time  a  lamp  is 
placed  on  the  photometer  have  been  grouped  together.  Some 
insignificant  changes  in  the  data  published  in  19 10  have  been 
made  to  correct  for  slight  errors  introduced  by  the  barograph 
then  used,  and  the  revised  results  are  given  in  Part  I  of  the 
table.     These  data  include  628  groups  of  sets,  or  about  30,000 

Water  Vapor  Correction  Factors. 

(I. — Observations  of  1910.) 

I,atnp                      Times  en  photometer  Factor  a  Weight 

Chance  1 16 74  0.00568  174 

Chance  118 23  0.00565  22 

Sugg  171 62  0.00565  130 

American  25  . 16  0.00584  11 

American  74 19  0.00564  48 

American  157 11  0.00581  8 

American  162 7  0.00572  9 

Weighted  mean  value  of  a  0.00567 

(II.— Observations  of  1911-1912.) 

Lamp  Times  on  photometer  Factor  a  Weight 

Chance  116 135  0.00586  226 

Chance  118 177  0.00552  339 

25  other  lamps 264  0.00569  269 

Weighted  mean  value  of  a  0.00567 

individual  settings  of  the  photometer.  The  data  obtained  from 
later  tests  are  summarized  in  Part  II  of  the  table ;  they  represent 

12  Transactions  Illuminating  Engineering  Society,  Vol.  V,  p.  766,  1910. 
18  loc.  cit. 


CRITTENDEN  AND  TAYLOR  I     PENTANE  LAMP  43 1 

about  2,300  groups  of  sets  or  perhaps  75,000  individual  photom- 
eter settings.  In  assigning  weights  to  individual  lamps  allowance 
was  made  both  for  the  number  of  times  on  the  photometer  and 
for  the  range  of  humidity  covered.  Since  ranges  as  small  as 
5  liters  were  included,  giving  a  total  variation  of  only  3  per  cent, 
in  these  cases,  and  some  of  the  lamps  were  on  the  photometer 
only  six  times,  the  results  from  individual  lamps  cannot  be  ex- 
pected to  agree  very  closely.  However,  only  two  lamps  gave  a 
value  for  a  below  0.0050  and  the  highest  value  obtained  was 
0.0066.  The  mean  result  checks  the  former  one  even  more  closely 
than  could  be  expected,  and  there  seems  to  be  no  room  for  doubt 
that,  at  least  in  the  Bureau  laboratory,  the  effect  is  very  definitely 
and  consistently  represented  by  a  factor  of  0.57  per  cent,  per  liter 
of  water  vapor.  This  factor  has  been  used  in  plotting  the  chart 
which  is  reproduced  in  Fig.  3. 

This  chart  has  been  so  plotted  that  the  departure  of  a  lamp 
from  the  normal  value  can  be  read  directly  from  it  when  the 
barometric  pressure  and  the  readings  of  the  wet  and  dry  bulb 
thermometers  of  a  ventilated  hygrometer  are  known.  The  hygro- 
meter may  be  either  a  sling  psychrometer  such  as  is  used  by  the 
U.  S.  Weather  Bureau  or  a  mechanically  ventilated  instrument 
like  the  Assmann  psychrometer.  If  an  ordinary  stationary  hygro- 
meter is  used  the  covering  of  the  wet  bulb  should  be  only  one 
thickness  of  very  thin  material  and  the  reading  should  be  taken 
at  the  lowest  point  to  which  it  can  be  brought  by  vigorous 
fanning. 

The  curves  at  the  left  of  the  chart  give  the  percentage  devia- 
tion from  normal  candle-power  which  corresponds  to  given 
temperature  and  wet  bulb  depression  when  the  barometric  pres- 
sure is  normal,  which  for  the  pentane  lamp  is  760  mm.  When 
the  pressure  is  different  from  this  the  additional  deviation  can 
be  found  from  the  curves  at  the  right  as  follows :  First  read  off 
the  deviation  in  the  regular  way  as  if  the  barometer  were  normal, 
and  note  this;  then  from  this  point  pass  horizontally  across  the 
sheet  to  that  curve  at  the  right  which  represents  the  actual  pres- 
sure. Vertically  above  or  below  the  point  at  which  this  curve 
is  reached  will  be  found  the  amount  to  be  added  to  the  deviation 
as  read  from  the  first  curves.     This  added  amount  is  made  up 


432 


TRANSACTIONS  I.   E.   S. PART  II 


chiefly  of  the  direct  effect  of  pressure  on  the  candle-power,  but 
it  includes  also  a  proper  allowance  for  the  fact  that  the  amount 
of  water  vapor  which  corresponds  to  a  given  pair  of  bulb  read- 
ings depends  somewhat  on  the  barometric  pressure. 

The  chart  is  plotted  for  the  pentane  lamp,  but  so  far  as  we  know- 
it  may  be  applied  to  other  flames  without  introducing  serious 
errors.  It  would  certainly  be  justifiable  to  measure  gas  flames. 
for  instance,  with  electric  standards  and  to  correct  the  observed 
candle-powers  to  obtain  normal  values  according  to  the  chart. 

In  order  to  give  some  idea  of  the  amount  of  variation  in 
water  vapor  which  occurs  from  season  to  season,  data  calculated 
from  the  records  of  the  Weather  Bureau  which  give  the  state  of 

~Z8 


SE4 


"16 


i 

fe4 


£ 

?/ 

\ 

& 

y/ 

\ 

<£ 

y 

/ 

* 

— - 

\ 

y 

A 

4: 

}y 

a^; 

3 

^ 

4 

& 

/y 

CHI 

:a« 

)-' 

*V 

A 

4/ 

ffc 

\ 

JP, 

v'/ 

,m  \  2 

& 

/. 

7? 

>, 

y/Z_ 

N 

V 

/, 

^ 

X 

^ 

5*. 

E^j 

JAN.        FEB.       MAR.      APR.        MAY      JUNE      JULY       AU6.        SEP        OCT.         NOV.        DEC 


Kig.  4. — Average  water-vapor  content  of  the  air  at  various  American 
cities  (1904  to  1908). 

the  out-door  air  have  been  plotted  in  Fig.  4.  In  general  the 
amount  of  moisture  is  small  in  winter  and  large  in  summer;  the 
average  for  May  and  for  October  is  practically  the  same  as  for 
the  whole  year.  The  variation  from  day  to  day  is  often  great, 
especially  in  the  spring  and  fall,  but  the  monthly  averages  show 
considerable  regularity  from  year  to  year.  The  curves  are  plot- 
ted from  the  averages  of  observations  for  5  years. 

The  normal  values  adopted  in  Europe  are  lower  than  the 
average  for  nearly  all  parts  of  the  United  States ;  consequently 
the  average  of  uncorrected  candle-powers  of  flame  standards  in 
this  country  runs  somewhat  lower  than  the  normal  intensities. 


CRITTENDEN  AND  TAYLOR  :     PENTANE  LAMP  433 

The  following  table  gives  the  average  amounts  of  water  vapor 
at  several  cities  for  the  period  of  5  years  plotted  in  Fig.  4.  The 
last  column  gives  the  corresponding  departure  of  a  pentane  lamp 
from  its  normal  value,  which  is  practically  the  percentage  of 
difference  between  the  nominal  and  the  actual  average  candle- 
power  of  any  source  which  is  measured  by  the  lamp. 

Average  Water-vapor  and  Pentane  Candle-power  in  the 
United  States. 

Per  cent,  departure  of 
Average  water-vapor  pentane  lamp  from 

Place  Liters  per  cm.  normal  value 

Boston    9.9  — 1.1 

Omaha 10.3  1.3 

Chicago 10.6  1.5 

Philadelphia 11. 2  1.8 

San  Francisco 11.3  1.9 

Washington    12.0  2.3 

New  Orleans 19. 1  6.3 

SUMMARY. 
In  this  paper  we  have  attempted  to  describe  fully  the  method 
of  testing  pentane  lamps  at  the  Bureau  of  Standards,  to  give 
brief  general  directions  for  the  use  of  the  lamps,  to  discuss  more 
fully  the  question  of  fuel,  and  to  furnish  data  on  the  effects  of 
pressure  and  moisture  in  a  form  which  may  be  useful  to  those 
who  have  occasion  to  reduce  observations  on  flames  to  normal 
candle-power  values.  Such  new  data  as  has  been  introduced  has 
been  obtained  in  the  laboratories  of  the  Bureau,  and  our 
acknowledgements  are  due  to  Dr.  Rosa,  under  whose  direction 
the  work  has  been  carried  on,  and  to  Mr.  G.  J.  Schladt,  who  has 
assisted  in  most  of  the  tests. 


DISCUSSION. 

Mr.  C.  O.  Bond  :  We  are  very  fortunate  indeed  to  have 
another  study,  or  a  continued  study,  of  the  pentane  lamp  from 
the  Bureau  of  Standards.  There  are  three  things  brought  out 
in  this  paper  that  I  would  hardly  have  suspected,  yet  I  have 
learned  that  when  the  Bureau  of  Standards  speaks,  it  is  danger- 
ous to  hold  opposite  opinions  on  the  subject  unless  extremely 
one  is  well  fortified,  because  they  investigate  very  thoroughly. 

On  the  ninth  page  it  is  stated  that  the  conviction  has  been 


434  TRANSACTIONS  I.   E.   S. — PART  II 

strengthened  "that  most  of  the  change  in  candle-power  caused  by- 
variation  in  conditions  of  operation  may  be  attributed  to  the  air 
circulating  system  and  the  variations  in  the  flow  of  air  which 
arise  from  changes  in  the  relative  temperatures  of  parts  of  the 
lamp."  In  the  laboratory  with  which  I  am  connected  we  have 
tried  to  study  the  effect  of  changing  the  temperatures  of  parts  of 
the  lamp  and  could  not  arrive  at  very  definite  formulae.  We  knew 
that  we  did  get  variations  in  lamp  values  depending  on  whether 
these  different  parts  were  changed  in  temperature.  For  instance, 
we  had  one  lamp  that  was  mounted  on  the  tripod  with  a  solid 
rod  supporting  it,  and  another  one  with  a  hollow  rod  supporting 
it,  and  we  got  two  different  values,  apparently  depending  on  the 
fact  that  one  rod  was  a  better  conductor  than  the  other  one.  By 
abstraction  of  heat  they  would  work  in  such  a  way  as  to  change 
the  flow  of  the  hot  air  which  leads  to  the  inner  part  of  the  flame, 
and  hot  air  determines  the  degree  to  which  the  pentane  flame 
shall  bulge.  In  this  way  the  air  changes  the  area  of  the  flame 
and  consequently  its  candle-power.  Where  a  booth  is  built 
around  a  lamp,  I  would  have  expected  to  see  some  change  in  the 
flow  of  air  at  least  as  supplied  to  the  flame  exterior  and  that  this 
would  have  an  effect  in  changing  the  candle-power  of  the  lamp. 

The  second  point  is  in  regard  to  the  change  in  pressure  in  the 
saturator  box.  We  were  fixed  in  our  belief  that  we  did  get  a 
considerable  change,  although  it  must  be  said  here  that  the  largest 
changes  came  when  we  were  using  a  regulating  valve  placed  near 
the  burner.  The  lamps  stand  some  thirty  inches  high  and  the 
regulator  cock  is  now  placed  high,  at  the  outlet  of  the  saturator 
box.  If  one  uses  the  regulator  near  where  the  vapor  tube  enters 
the  burner  that  is  a  somewhat  different  affair.  My  experience  was 
that  if  the  pressure  in  the  saturator  box  is  increased  and  then 
regulated  by  this  lower  valve,  when  the  vapor  passed  into  the 
burner,  it  had  a  higher  velocity,  and  that  the  effect  of  this  velocity 
had  not  entirely  died  away  before  the  issuance  of  the  vapor 
through  the  perforations  constituting  the  burner  tip ;  so  that 
there  was  apparently  a  stiffening  of  the  sides  of  the  flame;  the 
blue  part  of  the  flame  went  higher  with  the  added  pressure  and 
by  thus  decreasing  the  luminous  area,  the  candle-power  was 
lessened. 

The  saturator  box  was  certainly  intended  in  the  beginning,  as 


PENTANE  LAMP  435 

Dr.  Harcourt  has  said,  to  be  used  with  its  inlet  cock  open,  thus 
maintaining  a  constant  pressure  head.  I  have  always  gone  on 
the  assumption  that  with  a  constant  pressure  head,  with  uniform 
fuel  fed  through  a  constant  orifice,  there  would  be  a  constant 
consumption,  and  therefore  a  constant  flame,  leading  to  a  con- 
stant candle-power. 

The  third  point  which  is  of  great  interest  and  which  is  very 
important  indeed  is  the  quickness  with  which  they  claim  pentane 
as  now  furnished  to  the  market  seems  to  change  its  density. 
Mr.  Crittenden  has  said  that  after  one-tenth  of  it  has  been  used 
the  residue  is  already  well  beyond  the  limits  set  for  the 
density  of  the  pentane.  It  hardly  seems  possible  that  this  could 
have  been  true  at  the  time  Dr.  Harcourt  invented  his  lamp  and 
made  the  specification  for  its  use.  I  am  wondering  whether  it 
is  not  now  possible  to  meet  the  tests  provided  in  the  specifica- 
tions and  yet  not  supply  the  exact  article  that  was  intended  in 
the  original  specifications.  If  I  remember  aright,  the  specifica- 
tion of  light  American  petroleum  such  as  is  known  as  gasoline 
from  which  to  refine  the  pentane,  referred  originally  to  Pennsyl- 
vania oil  with  a  paraffin  base.  This  is  not  so  readily  supplied 
now  as  formerly,  and  it  may  be  due  to  this  lack  of  a  paraffin 
base  that  the  Bureau  has  found  such  a  rapid  change  in  the  pen- 
tane density.  I  do  not  claim  any  knowledge  in  this  direction 
(but  the  matter  is  of  such  importance  to  the  reputation  of  the 
pentane  lamp  as  a  standard  of  reference,  that  the  subject  should 
be  investigated  further,  and  access  be  had  to  original  records  in 
England  to  find  where  this  discrepancy  has  crept  in). 

Mr.  C.  W.  Jordan  :  I  would  like  to  inquire  in  regards  to  the 
relative  accuracy  of  carbon  dioxid  determinations  in  air  with  a 
Zeiss  refractometer  and  by  chemical  methods.  It  occurred  to 
me  that  the  refractometer  shows  its  greatest  accuracy  when  only 
two  constituents  of  a  gaseous  mixture  are  variable.  By  Petters- 
son's  chemical  method,  carbon  dioxid  in  air  can  be  determined 
accurately  to  within  0.002  per  cent. 

Have  the  authors  of  this  paper  made'  any  extensive  investiga- 
tions of  pentane  from  a  chemical  standpoint?  While  the  fuel 
consumed  in  the  lamp  is  designated  by  the  formula  Cr,H12  it  is 
admittedly  not  pure  pentane.     In   fact  the  portions  distill  off 


436  TRANSACTIONS  I.   E.   S. PART  II 

between  25  °  C.  and  400  C.  which  precludes  a  definite  compound. 
Prepared  from  the  rather  vaguely  defined  American  petroleum, 
it  is  said  to  consist  chiefly  of  pentane  with  small  quantities  of 
lower  and  higher  homologues  whose  presence  does  not  affect, 
seriously,  the  candle-power  of  the  lamp.  Pentane  is  the  fifth  of 
the  paraffin  series  of  hydrocarbons,  the  first  four  of  which, 
methane,  ethane,  propane  and  butane  being  gaseous  at  ordinary 
temperatures.  Methane  burns  with  a  non-luminous  flame  and 
the  candle-power  of  the  succeeding  compounds  increase  pro- 
gressively, due  to  the  increase  in  the  percentage  of  free  carbon 
dissociated.  Pentane  is  known  to  exist  in  three  modifications, 
each  having  different  boiling  points,  and  the  separation  of  admix- 
tures of  propane,  butane  and  the  higher  homologue,  hexane,  is 
very  difficult. 

For  lack  of  evidence  to  the  contrary,  I  am  inclined  to  believe 
that  serious  variations  in  candle-power  may  occur  in  different 
lots  of  pentane  having  the  proper  specific  gravity  and  giving 
negative  tests  for  the  hydrocarbons  of  the  benzene  series.  I 
think  that  flame  standards  should  consume  a  definite,  chemically 
pure  fuel  and  in  the  case  of  the  Harcourt  pentane  lamp,  syn- 
thetically produced  pentane  would  be  the  ideal  fuel. 

On  the  fourth  page  of  this  paper  a  statement  is  made  that  pen- 
tane lamps,  manufactured  by  the  same  maker  and  supposedly  of 
the  same  type,  sometimes  show  variations  in  candle-power  of  as 
great  as  4  per  cent.  This  variation  is  believed  to  be  due  to 
mechanical  differences  in  the  lamps.  While  it  is  exceedingly 
important  to  standardize  lamps  under  the  existing  conditions,  I 
think  that  the  line  of  future  investigation  should  be  that  of  draw- 
ing up  standard  specifications  to  eliminate  these  differences. 

Mr.  E.  C.  Crittenden  (in  reply)  :  Mr.  Bond's  complimentary 
remarks  regarding  the  work  done  on  pentane  lamps  at  the  Bureau 
of  Standards  are  very  gratifying.  It  is  a  pleasure  also  to 
acknowledge  our  indebtedness  to  Mr.  Bond  for  valuable  sugges- 
tions which  he  has  given  us  at  various  times  during  the  progress 
of  this  work.  The  position  which  the  pentane  lamp  now  holds 
in  this  country  is  due  in  no  small  degree  to  the  pioneer  work 
which  he  did  in  introducing  it,  and  his  thorough  personal  knowl- 
edge of  the  lamp  gives  much  weight  to  his  suggestions. 


PENTANE  LAMP  437 

It  has  been  stated  that  the  relative  temperatures  of  various 
parts  of  the  air  circulating  system  of  the  lamp  are  probably  the 
cause  of  most  of  the  changes  in  candle-power  which  occur  when 
the  lamp  is  operated  under  different  conditions.  The  case  in 
which  a  change  of  candle-power  was  brought  about  by  substi- 
tuting a  solid  rod  for  the  usual  hollow  one  at  the  base  of  the 
lamp  is  an  interesting  example,  which  would  be  hard  to  explain 
in  any  other  way.  Since  the  temperature  may  be  appreciably 
affected  by  apparently  unimportant  changes  in  the  conditions  of 
operation,  it  is  evident  that  if  the  lamp  is  to  be  enclosed,  great 
care  must  be  taken  to  make  sure  that  the  enclosure  does  not 
affect  the  candle-power.  As  is  stated  in  the  paper,  it  is  better  not 
to  use  any  screening  except  that  necessary  to  shut  out  stray  light. 
The  additional  screens  described  were  used  because  of  unfavor- 
able conditions  in  the  laboratory,  which  would  otherwise  have 
prevented  accurate  measurements  at  times  when  drafts  were  bad ; 
the  particular  screen  described  was  found  by  comparative  meas- 
urements to  give  no  perceptible  effect  on  the  candle-power. 

With  regard  to  the  change  in  density  of  the  pentane  and  the 
possibility  that  the  pentane  commercially  supplied  at  present  is 
not  the  same  as  when  the  standard  was  adopted  by  the  Gas 
Referees,  because  the  source  of  supply  is  different,  I  would  say 
that  if  the  material  is  prepared  according  to  the  Referees'  direc- 
tions the  products  obtained  from  various  kinds  of  crude  oil  ought 
not  to  differ  materially.  Chemists  who  have  studied  various  oils 
say  that  the  differences  between  them  are  largely  in  the  heavier 
constituents ;  so  far  as  the  very  light  fractions  are  concerned, 
Western  oils  are  not  essentially  different  from  Pennsylvania  oil. 
In  the  paraffin  series  the  next  substance  above  pentane  is  hexane, 
and  that  form  of  it  which  appears  in  petroleum  has  a  boiling 
point  of  about  6o°  C.  The  four  distillations  specified  below  that 
temperature  ought  to  remove  the  hexane  quite  completely.  The 
final  distillation  is  carried  over  the  rather  wide  range  from  25  ° 
to  400  because  the  pentane  itself  exists  in  two  forms  having 
boiling  points  about  280  and  360  C.  Incidentally,  the  densities 
of  the  two  forms  are  0.625  and  0.631,  and  it  is  not  clear  why  the 
referees  set  the  narrow  limits  of  0.6235  to  0.626  in  the  specifi- 
cations. 

There  is  little  danger  of  errors  arising  from  failure  to  remove 
5 


43§  TRANSACTIONS  I.  E.   S. — PART  II 

the  other  members  of  the  paraffin  series,  but  amylene  (C5H10), 
of  the  define  series,  cannot  be  separated  from  the  pentane  by 
distillation  alone,  because  it  has  several  forms  whose  boiling 
points  come  between  25 °  and  400  C.  The  sulphuric  acid  treat- 
ment prescribed  should,  however,  remove  it;  the  permanganate 
test  is  supposed  to  show  whether  this  has  been  properly  done. 

It  is  by  no  means  certain  that  the  pentane  on  the  market  is 
prepared  in  accordance  with  the  Referees'  directions.  To  make 
up  a  mixture  which  shall  have  the  required  density  is  not  diffi- 
cult, and  consequently  it  is  desirable  that  the  chemical  test  be 
made. 

By  repeated  distillation,  or  by  synthesis,  it  is  possible  to  obtain 
practically  pure  pentane,  but  the  cost  would  prohibit  its  use  in 
testing.  If  it  were  a  matter  of  setting  up  a  primary  standard 
the  pure  fuel  might  be  prepared.  There  are,  however,  other 
difficulties  to  be  met  before  the  lamp  can  be  seriously  considered 
as  a  primary  standard;  some  time  it  may  be  so  perfectly  under- 
stood that  we  shall  be  willing  to  base  the  unit  upon  it,  but  per- 
sonally I  would  say  that  our  hopes  of  attaining  that  end  are  not 
high. 

T.  J.  Little,  Jr.  :  I  notice  on  the  twenty-third  page  it  is  stated 
that  "The  chart  is  plotted  for  the  pentane  lamp,  but  so  far  as  we 
know  it  may  be  applied  to  other  forms  without  introducing  serious 
errors."  I  suppose  that  is  meant  for  the  naked  flames,  as  for  in- 
stance, flat  flame  gas  burners,  but  in  the  accurate  determination 
for  candle-power  of  incandescent  mantle  burners  such  as  those 
for  use  on  either  gas  or  gasoline,  would  you  also  recommend  cer- 
tain correcting:  factors? 


KINGSBURY:     SUNDAY-SCHOOL   ROOM    ILLUMINATION         439 

EXPERIMENTS  IN  THE  ILLUMINATION  OF  A 
SUNDAY-SCHOOL  ROOM  WITH  GAS.* 


BY    EDWIN    F.    KINGSBURY. 


Synopsis:  The  author  describes  the  lighting  by  gas  of  a  typical 
Sunday-school  room,  composed  of  a  large  central  floor  with  a  high  glass 
paneled  ceiling  and  alcoves  at  the  ends.  Four  4-burner  "arcs"  having  a 
magnet-valve  pilot  system  of  control  from  the  main  floor  were  placed 
above  the  ceiling  to  light  the  central  area.  The  alcoves  were  lighted  from 
the  side  walls  by  placing  a  large  semi-circular  translucent  screen  in 
front  of  each  of  the  eleven  upright  burners.  The  object  of  these  screens 
was  to  transmit  part  of  the  light  directly  and  to  reflect  a  portion  back 
on  to  the  wall.  The  whole  forms  a  light  source  of  large  area  and  low 
intrinsic  brilliancy. 

When  planning  the  lighting  of  a  church,  the  auditorium  is 
usually  carefully  studied  from  every  angle  to  secure  ample,  uni- 
form illumination  and  to  bring  out  the  best  artistic  effect.  By 
the  time  the  Sunday-school  room  is  reached  ideas  of  economy 
become  strong  and  almost  any  style  of  illumination  is  made  to  do, 
though  this  room  may  actually  be  used  more  frequently  and  for 
more  varied  purposes  than  the  auditorium.  While,  from  the 
nature  of  the  two  rooms,  it  may  not  be  desirable  to  aim  for  the 
same  effect  in  the  Sunday-school  room  as  in  the  church  proper, 
still  the  lighting  of  the  former  should  be  carefully  studied  to  in- 
clude the  best  principles  of  a  good  practical  installation. 

The  purpose  of  this  paper  is  to  describe  some  experiments  in 
the  illumination  of  the  Sunday-school  room  of  the  Summit  Pres- 
byterian Church,  of  Germantown,  Philadelphia. 

This  room  is  typical  of  many  used  for  Sunday-school  pur- 
poses, being  composed  of  a  large,  open  central  portion  with 
alcoves  at  two  sides  which  can  be  closed  and  utilized  for  classes. 

The  lighting  requirements  for  such  a  place  demand  that  the 
sides  shall  be  treated  more  as  distinctive  rooms  and  that  two 
separate  plans  of  lighting  be  adopted.    The  most  usual  solution 

*A  paner  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


440  TRANSACTIONS  I.  E.   S. — PART  II 

provides  one  or  more  chandeliers  over  the  main  floor  and  a 
smaller  one  in  the  center  of  each  alcove.  In  the  present  case 
there  was  a  stained  glass  ceiling  over  the  center  which  could  b'e 
used  advantageously  and  in  each  alcove  there  were  two  wall  out- 
lets that  it  was  desirable  to  use. 

There  were  several  reasons  for  placing  the  light  sources  above 
the  ceiling  in  the  central  portion.  First,  it  removed  the  light  from 
the  field  of  vision  and,  secondly,  it  utilized  the  beauty  of  the 
stained  glass,  which  any  lamps  hanging  below  would  tend  to  ob- 
scure. Then,  also,  the  attic  was  easy  of  access  for  maintenance 
purposes. 

The  problem  of  lighting  the  sides  was  a  more  difficult  one  to 
solve,  as,  utilizing  the  present  outlets,  brought  the  lights  too  low, 
especially  as  an  audience  sits  where  the  light  from  one  row  of 
lamps  would  be  square  in  the  eyes.  Even  well  frosted  globes  were 
objectionable,  as  the  walls  were  dark  and  the  contrast  strong. 

In  accordance  with  what  has  been  said  the  work  may  conven- 
iently be  divided  into  two  portions — the  first  dealing  with  the 
permanent  lighting  of  the  central  portion  of  the  room  through 
the  stained  glass  ceiling  and  the  second  with  a  temporary  instal- 
lation along  the  front  and  rear  to  illuminate  the  alcoves. 

The  plan  of  the  room,  with  essential  dimensions,  is  shown  in 
Fig.  i.  The  room  is  60  ft.  (18.29  m-)  l°ng  by  35  ft.  (10.67  m-) 
wide  and  35  ft.  (10.67  m-)  high.  Twelve  feet  (3.66  m.)  from  the 
front  and  back  walls  open  archways  rise,  the  walls  in  front  at 
the  top  curving  inwardly  to  meet  a  glass  paneled  ceiling  28  ft. 
(8.53  m.)  long  by  42  ft.  (12.80  m.)  wide,  each  panel  being  3  ft. 
by  3  ft.  (0.91  m.  x  0.91  m.).  Heavy  blinds  slide  between  the  arch 
posts  at  the  rear  to  form  three  separate  rooms.  The  wall  north 
of  the  main  floor,  composed  of  three  heavy  Venetian  blinds  which 
can  be  raised  to  include  the  low-ceiling  room  beyond  it  as  a  part 
of  the  school  room,  rises  to  only  one-third  of  the  full  height  of 
the  room,  the  space  above  being  entirely  open.  These  three  blinds, 
however,  were  closed  throughout  the  test.  At  the  south  are  large 
double  sliding  doors  in  the  center,  the  body  of  the  wall  being 
plain  and  curving  inwardly  at  the  top  to  meet  the  ceiling,  as  in  the 
front  and  rear.     Wooden  wainscoating  5  ft.   (1.52  m.)  high  ex- 


KINGSBURY:     SUNDAY-SCHOOL   ROOM    ILLUMINATION         44I 

tends  around  the  room,  the  north  side  excepted.  All  wood  work 
is  dark  oak  in  color  and  the  walls  dark  green. 

The  room  is  lighted  in  the  daytime  by  three  large  stained  glass 
windows  at  the  rear.  The  three  windows  in  front  are  now  of 
no  use,  being  covered  by  a  later  built  portion  of  the  church. 

The  body  of  the  room  was  originally  lighted  by  12  small  in- 
verted incandescent  gas  burners  at  each  end,  distributed  uni- 
formly along  each  horizontal  beam  15  ft.  (4.57  m.)  high  at  the 
base  of  the  arches  and  supplied  with  pilots  and  pendant  chains  for 
lighting. 


SLIDING    DOORS  I..    '.'.     '  1 

Fig  i.— Plan  of  room. 


The  illumination  of  the  alcoves  was  provided  by  three  small 
inverted  units  and  several  open  flame  burners.  This  installation 
was  unsatisfactory,  for  at  least  two  reasons.  First,  the  mechani- 
cal operation  of  reaching  up  with  a  hook  to  catch  24  rings  was 
both  monotonous  and  time-consuming,  especially  if  the  room  was 
rather  dark  at  the  start,  or  if  many  rings  were  accidentally  set 
to  swinging  before  being  caught.  Secondly,  the  glare  from  the 
row  at  one  end  in  the  eyes  of  a  person  sitting  at  the  other  end  was 
annoying. 

These  two  rows  of  units  lighting  the  body  of  the  room  were, 


442  TRANSACTIONS  I.  E.   S. — PART  II 

therefore,  removed  and  the  glass  paneled  ceiling  utilized  by  plac- 
ing a  four-burner  inverted  incandescent  gas  arc  lamp  above  each 
of  four  panels,  as  indicated  by  the  circles  in  Fig.  i.  The  magnet- 
valve  pilot  system  of  ignition  is  used.  Two  push  buttons  are  lo- 
cated on  the  main  floor  near  the  southeast  door ;  each  one  lights  a 
pair  of  lamps.  Each  lamp  is  equipped  with  a  polished  aluminum 
conical  focusing  reflector  n  in.  (27.94  cm.)  diameter  at  the  top, 
26  in.  (66  cm.)  at  the  bottom  and  16  in.  (40.64  cm.)  high.  The 
mantles  are  30  in.  (76.2  cm.)  from  the  glass  ceiling  and  the  bot- 
tom of  the  reflector  16  in.  (40.64  cm.).  Opal  globes  were  decided 
upon  as  the  most  suitable.  Clear  globes  gave  a  bright  spot  in  the 
center  with  bright  concentric  rings.  The  spots  were  not  visible 
without  the  effort  to  look  up  at  a  high  angle  at  the  lights,  but  the 
illumination  on  the  floor  below  any  lamp  was  nearly  double  that 
half-way  between.  Opal  globes,  however,  obviated  this  difficulty, 
as  shown  in  Figs.  4  and  5.  This  system  has  been  in  use  a  year 
and  a  half  and  has  given  satisfaction. 

The  desultory  lighting  in  the  rear  of  the  arches  was  untouched 
until  recently,  when  the  experiment  described  below  was  tried. 
This  consisted  of  placing  upright  incandescent  gas  burners  on  the 
six  outlets  in  the  front  and  on  five  of  the  six  in  the  rear.  Ten 
inches  (25.4  cm.)  in  front  of  each  lamp  was  a  semi-circular  trans- 
lucent paper  screen,  highly  glazed  on  the  inner  side  and  matt  on 
the  outer.  A  plain  creamy-white  bathroom  wall  paper  answered 
the  temporary  purpose  very  well.  The  object  of  this  was  to  have 
the  glazed  surface  reflect  part  of  the  light  to  the  wall.  Then  the 
light  reflected  from  the  wall,  with  that  fraction  transmitted  by  the 
screen,  is  supposed  to  form  a  continuous  light  source  of  large 
area  and  low  intrinsic  brilliancy.  The  position  of  the  curve  is 
imitative  of  daylight  window  illumination  and  should  partake 
of  the  naturalness  of  the  latter.  If  the  wall  is  ornamental  the 
screens  may  well  be.  In  the  present  case  both  were  left  plain. 
The  screen,  ten  inches  (25.4  cm.)  high  with  a  radius  of  ten 
inches  (25.4  cm.)  and  having  the  light  at  the  center,  shields  the 
source  from  practically  every  position  a  person  would  be  likely 
to  assume.  The  exception  is  that  when  one  sits  sideways  and 
almost  directly  beneath  a  lamp  the  bare  mantle  can  be  seen  by 


KINGSBURY!     SUNDAY-SCHOOL   ROOM    ILLUMINATION         443 

looking  upward.  This  can  easily  be  remedied  by  making  the 
screen  a  little  wider  or  lowering  the  present  one  somewhat. 

In  order  to  prevent  too  much  light  being  thrown  back  against 
the  wall  and  to  direct  more  obliquely  to  the  floor,  as  sunlight 
would  be,  a  small  rectangular  piece  of  bright  tin  was  placed 
(experimentally)  directly  in  the  rear  of  each  mantle,  allowing  it 
to  slant  forward  a  trifle.  By  selecting  the  proper  width  and  angle 
for  such  a  reflector  any  desired  effect  in  this  connection  could  be 
obtained. 

Owing  to  the  high  absorption  of  the  wall  and  woodwork,  the 
employment  of  a  more  efficient  diffusive  surface  back  of  each  unit 
was  imperative  to  secure  sufficient  illumination.  For  this  pur- 
pose a  semi-matt  white  paper  36  in.  by  36  in.  (0.913  x  0.913  m.) 
was  pasted  on  the  wall. 

Fig.  2  shows  the  front  of  the  room  and  Fig.  3  the  rear  under 
this  artificial  illumination.  In  the  former  can  be  seen  the  light 
reflected  from  the  floor,  which  appears  as  a  bright  band  running 
parallel  with  the  wall.  It  will  be  noted  at  the  top  of  the  picture, 
above  the  horizontal  beam,  that  there  is  considerable  light  thrown 
obliquely  upwards  and  a  design  painted  on  the  under  side  of  the 
arches  was  brought  out  conspicuously  clear.  Unfortunately  the 
wall  near  the  ceiling  could  not  be  changed,  but  if  it  were  white  or 
a  light  cream  it  would  help  out  in  a  line  where  most  needed — 
the  transition  from  the  side  to  the  main  overhead  illumination. 

The  uniformity  of  the  screens  and  background  is  shown  in 
both  pictures  where  they  cannot  be  distinguished,  except  at  an 
oblique  angle,  as  in  Fig.  3.  This  system  seems  to  solve  success- 
fully the  illumination  of  the  alcoves. 

In  Fig.  6  is  shown  a  candle-power  distribution  curve  of  one 
of  the  wall  units,  taken  through  the  plane  bisecting  the  central 
angle  of  1800  subtended  by  the  screen.  Each  arm  of  the 
curve  has  its  own  particular  function.  The  two  back  loops  go  to 
the  wall  to  be  reflected,  one  to  the  floor  and  the  other  at  an 
oblique  angle  toward  the  ceiling,  while  a  considerable  part  of 
both  will  be  diffused  into  the  space  in  front  of  the  unit.  The 
upper  front  one  goes  to  the  ceiling  and  the  lower  front  one  to 
the  floor.    It  is  desirable  to  have  the  curve  directly  in  the  rear  of 


444  TRANSACTIONS  I.   E.   S. PART  II 

the  screen  swing  in  more,  as  the  light  thrown  straight  back  is 
more  or  less  useless.  This  can  be  partially  accomplished  by  mak- 
ing the  tin  reflector  larger,  or  by  having  the  semi-circular  screen 
not  as  a  section  cut  from  a  cylinder,  but  as  a  section  of  an  hour- 
glass. Even  then  any  diffused  light  would  keep  the  wall  quite 
bright. 

Illumination  measurements  were  taken  with  all  the  room  fully 
lighted,  and  with  the  center  lamps  and  the  lamps  at  one  end 
lighted  separately,  the  object  being  to  show  the  effect  of  the  two 
distinct  systems  on  each  other.  All  readings  were  taken  in  a  30 
in.  (76.2  cm.)  plane,  with  the  exception  of  the  readings  on  the 
platform,  which  were  made  on  a  plane  2  ft.  (60.96  cm.)  higher 
than  those  on  the  floor. 

Fig.  4  shows  the  first  case  when  all  the  lamps  are  at  maximum 
brilliancy.  This  seems  to  give  the  best  effect  and  with  the  light 
properly  spread  no  fatigue  or  annoyance  should  be  experienced 
by  an  audience  facing  them  for  several  hours.  However,  the 
alcove  lighting  is  very  elastic,  and  if  the  maximum  light  desired 
at  one  time  is  considered  excessive  at  another  time,  the  lamps  can 
easily  be  turned  down,  as  they  are  upright  burners  and  will  work 
equally  well  at  any  intensity. 

An  analysis  of  the  results  shows  a  high  candle-power  close  to 
the  lamps,  with  a  rapid  falling  off  to  a  minimum  directly  beneath 
the  arches.  Then  comes  a  gradual  rise  to  a  maximum  on  the 
main  floor. 

Fig.  5  shows  the  center  and  one  end  lighted  separately.  As 
will  be  noted  the  end  wall  illumination  has  an  appreciable  effect 
on  the  main  floor,  the  result  being  to  enlarge  the  contours.  It 
would  be  more  desirable  if  the  ends  sent  more  light  beneath  the 
arches,  and  this  could  easily  be  secured  by  making  the  entire 
wall  much  lighter. 

The  average  foot-candles  on  the  30  in.  (76.2  cm.)  plane  in  the 
center  is  1.08.  While  this  may  seem  low,  it  has  proven  to  be 
ample  for  all  purposes  and  there  seems  to  be  a  peculiar  advantage 
in  reading  responses  or  singing  where  the  book  is  held  fairly  hori- 
zontal. The  vertically  directed  light  here  from  a  height  gives  one 
a  feeling  the  same  as  that  experienced  in  cathedrals  and  is  hence 


l^ 


Fig.  2. — Front  of  room  lighted. 


y 

L  k 

•f   ! 

n 

^1 

^'"i*"*— r; 

— 

Is    - 

r 

^''oWi* 

1          * 

hliMfl 

■.■■-«^. 

Fig.  3.  — Rear  of  room  lighted. 


«0  J    'tlS 


~~i7~Jrri;     iar^~n 


fc- 

/^?0 OS 7 


^V7\  MOB: 
-   »  f  -  S^ODE 

\  '        A Nl >■'»-,—  ->J- . < 

>X  \     \  (      \  -<^„ 

■/>  -)l-9i        zis^        '  '061         \  aw     v     ~" ~ai "" " 

^b       W777ZW///////////////A 


x i^DDidQft] 

"DTsoiio 

KliQJDiDlnTaa 


E^ 


1       W 

\ s'975       W0\      v      400 


W   ^ 


tit 


PLANE  OF  READIN6S  30' HIGH 

Fig.  4. — Illumination  readings. 


ALL  LIGHTS  OK  FULL 


.~)-1.80     I     1.60 

!   [oh 


4.00   \  [ahe       a30 


□HHBQi 

oBy'SB 

01 


0 


^/HTToo       £I5>      J> 


isn 


so 
Op 


10000 


-#- 


N 

•~}2.90     '   1.70      v'    C.50 


-~  J.jJ0_ 
0.85  0.95 


<1 
■ 


I 


'i  !       , 


PLANE  OF  READING  30"HI6H.     END  &CENTER  LIGHTS  ON  SEPARATELY  AS  MARKED. 
Fig.  5. — Illumination  readings. 


KINGSBURY:     SUNDAY-SCHOOL   ROOM    ILLUMINATION 


445 


appropriate.  Experience  has  shown  the  body  of  the  room  is  suf- 
ficiently well  lighted  by  the  four  arc  lamps,  but  they  have 
little  influence  back  of  the  arches. 

There  are  spots  under  the  archways  where  the  illumination  is 
a  trifle  too  low  to  allow  of  continued  reading,  but  these  would  be 
corrected  by  lighter  side  walls  in  a  finished  installation. 

This  installation  is  described  with  the  purpose  of  showing  that 
in  rooms  so  cut  up  and  apparently  hard  to  illuminate  well,  the 
very  peculiarities  may  be  used  to  provide  something  at  once  un- 
usual and  practical.  Certainly,  the  illuminating  engineer  should 
not  feel  that  any  situation  is  so  commonplace  and  unimportant 


r   /\/   y\  ^-V-" 

— *"f»v  7\   \/\   s 

80    I    60    [   40    |    V  fl 

y^ZO    |40-4v60        8.0 

XXfRont^^/^ 

_^J\r-^^ACK/<C/      , 

Fig.  6. — Candle-power  distribution  through  vertical 
plane  of  an  alcove  unit. 

that  any  casual  style  is  good  enough,  but  he  should  be  alive  to 
every  opportunity  to  make  the  uncommon  out  of  the  common. 
This  means  that  while  there  are  certain  fundamental  rules  always 
to  be  observed,  they  make  only  a  beginning,  and  the  remainder 
of  his  task  depends  on  his  originality.  This  is  not  assuming,  of 
course,  that  expense  is  to  be  overlooked.  In  the  case  described 
above  a  relatively  costly  installation  was  a  poor  one.  It  means 
that  thought  and  study  must  be  given  to  each  individual  case. 

The  thanks  of  the  author  are  due  to  Mr.  Charles  O.  Bond,  Dr, 
Herbert  E.  Ives  and  Mr.  C.  W.  Jordan  for  assistance  in  prepar- 
ing this  paper. 


44^  TRANSACTIONS  I.  E.   S. — PART  II 

DISCUSSION. 

Mr.  J.  D.  Israel:  I  feel  privileged  to  discuss  the  question 
of  church  lighting  from  a  commercial  standpoint.  The  psycho- 
logical effect  of  lighting,  discussed  by  Dr.  Lewis*  yesterday, 
I  think  should  be  considered  when  one  studies  church  lighting 
problems  from  a  purely  commercial  standpoint. 

I  wish  to  say  for  your  information  that  the  Philadelphia 
Electric  Company  approaches  this  problem  purely  on  a  com- 
mercial basis,  the  same  as  they  go  after  business  in  any  other 
class  of  lighting.  We  address  letters  to  the  trustees  and  pas- 
tors of  the  various  churches;  we  circularize  and  give  illustra- 
tions of  such  installations  as  we  already  have  connected  to  our 
circuits.  Then  we  follow  with  personal  appeals;  our  men 
appear  before  the  trustees.  We  ask  the  pastor  and  trustees  to 
consider  the  matter  as  a  commercial  problem,  as  an  aesthetic 
problem  and  as  a  religious  problem.  We  accomplish  good 
results  without  detracting  at  all  from  the  sacredness  of  the 
work. 

Mr.  E.  F.  Kingsbury:  One  use  of  the  alcove  lighting  not 
mentioned  in  the  paper  is  in  the  illumination  of  small  rooms, 
especially  in  residences,  where  there  are  wall  outlets  that  one 
dislikes  to  use  on  account  of  the  difficulty  in  avoiding  objection- 
able glare  from  them.  With  a  little  care  the  light  thrown 
downward  can  be  utilized  for  reading  on  a  table  near  the  wall 
and  the  light  thrown  upward  will  supply  the  general  illumina- 
tion. The  success  of  such  an  installation  depends  on  keeping 
the  brilliancy  of  the  screen  and  reflecting  walls  low.  This  is 
well  secured  on  the  screen  and  an  artistic  end  realized  if  it  is 
made  ornamental.  In  this  way  an  almost  useless  outlet  might 
be  turned  into  a  thing  of  beauty. 

*  "The  Psychic  Values  of  Light,  Shade,  Form  and  Color,"  Trans.  I.  E.  S.,  p.  357 
(October,  1913). 


LANSINGH  :     ENCLOSING   GLASSWARE  447 

CHARACTERISTICS  OF  ENCLOSING  GLASSWARE.* 


BY  VAN   RENSSELAER   LANSINGH. 


Synopsis:  In  this  paper,  enclosing  glassware  is  divided  into  two 
classes :  one,  purely  transmitting  and  diffusing,  such  as  ground,  opal  and 
leaded  glass;  the  other,  prismatic  glass,  which  employs  the  principle  of 
specular  reflection.  Photometric  curves  and  data  are  given  to  show  (r) 
that  with  glassware  of  the  first  class,  little  except  good  diffusion  and 
low  absorption  may  be  expected,  the  re-distribution  of  light  being  negli- 
gible; (2)  with  prismatic  glass,  the  distribution  may  be  varied  in  accord- 
ance with  the  wishes  of  the  engineer.  The  absorption  in  both  given  classes 
is  about  the  same. 

The  two  most  pronounced  tendencies  of  the  day  in  interior 
illumination,  excluding  the  industrial  field,  are,  first,  the  use  of 
indirect  and  semi-indirect  lighting,  and  second,  enclosing  glass- 
ware. By  the  latter  is  meant  the  use  of  large  diffusing  glass- 
ware, completely  surrounding  the  lamp.  A  great  deal  has  been 
written  in  the  technical  press  and  elsewhere  on  the  first  class, 
but  little  has  appeared  regarding  the  second.  The  present 
paper,  therefore,  aims  to  state  briefly  some  of  the  chief  character- 
istics of  this  class  of  lighting  units.  So  far  the  writer  has  made 
a  study  only  with  tungsten-filament  lamps,  but  with  the  possible 
exception  of  some  changes  in  color  characteristics,  it  is  believed 
the  same  results  would  be  obtained  with  gas  mantle  burners. 
This  cannot  be  said  of  arc  lamps,  however,  when  the  position  of 
the  arc  travels,  and  no  attempt  is  made  here  to  cover  this  phase 
of  the  subject. 

The  chief  characteristics  of  enclosing  glassware  are:  (1)  dis- 
tribution, (2)  absorption,  (3)  appearance,  (4)  effect  on  the  eye, 
(5)  effect  on  the  lamp,  and  (6)  color. 

The  question  of  appearance,  color,  and  effect  on  the  eye  are 
not  taken  up  here,  as  these  subjects  have  been  covered  by 
numerous  writers  in  the  Transactions  and  elsewhere.  The  effect 
on  the  lamp  need  not  be  given  much  consideration  as  a  number 
of  tests  have  shown  that  the  rise  in  temperature,  due  to  enclosing 
the  lamp,  is  not  sufficient  to  affect  its'  life. 

The  tests  reported  in  this  paper  were  made  at  three  different 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


448  TRANSACTIONS  I.   E.   S. — PART  II 

laboratories,  whose  methods  and  results  have  been  carefully 
checked  against  each  other ;  so  that  the  curves  given  are  all  com- 
parable. Inasmuch  as  comparative  rather  than  absolute  values 
are  desired,  the  curves  are  given  without  the  actual  candle- 
power.  All  curves,  however,  are  plotted  on  the  basis  of  1,000 
lumens  for  the  bare  lamp,  i.  e.,  the  candle-power  actually  found 
at  every  reading  in  the  tests  is  multiplied  by  the  ratio  of  1,000 
to  the  actual  lumens  of  the  lamp.  Thus,  in  the  case  of  a  100- 
watt  lamp  giving  908  lumens,  the  candle-power  readings  would 

be  multiplied  by  ■   '         =1.1.     By  thus  reducing  all  the  curves 

to  the  basis  of  a  1,000  lumen  lamp,  the  curves  can  be  compared 
without  reference  to  the  size  of  lamps  used.  For  example,  it 
will  be  noted  that  the  bare  lamp  curve  is  the  same  in  every  case, 
irrespective  of  the  size  of  lamp  tested.  In  every  case,  the  fig- 
ures given  are  based  on  lamp  flux  rather  than  emitted  flux,  as  the 
engineer  is  really  concerned  with  the  actual  flux  of  a  lamp  which 
is  available  for  any  given  purpose,  rather  than  the  relative 
emitted  flux  in  different  zones. 

There  are  many  different  kinds  of  enclosing  glassware  on  the 
market  at  the  present  time,  and  typical  examples  from  the  dif- 
ferent classes  were  selected  for  the  purpose  of  the  above  men- 
tioned tests.    The  kinds  of  glassware  selected  were  as  follows : 

Class  I — Pressed  opal  ball  in  two  pieces. 

Class  II — Blown  opal  ball. 

Class  III — Blown  opal  acorn. 

Class  IV — Cased  opal  ball. 

Class  V — Leaded  opal  ball. 

Class  VI — Ground  glass  ball. 

Class  VII — Prismatic  deep  reflector-bowls. 

Class  VIII — Prismatic  shallow  reflector-bowls. 

Class  IX — Prismatic  reflector-balls. 

The  actual  trade  names,  designations,  etc.,  of  the  glassware 
tested,  and  full  data  on  each  test,  are  given  in  an  appendix  to 
this  paper,  for  those  who  wish  this  information.  The  accom- 
panying illustrations  will  serve  to  identify  the  general  appearance, 
contour,  etc.,  of  each  unit. 


LANSINGH  :     ENCLOSING   GLASSWARE)  449 

Fig.  I  shows  a  two-piece  pressed  opal  ball  of  comparatively 
light  density,  listed  above  as  Class  I.  Fig.  2  shows  the  curve 
of  the  bare  lamp  and  a  10-inch  (25.4  cm.)  ball  tested  with  a 
100-watt  tungsten  filament  lamp.  Fig.  3  shows  the  distribution 
from  a  14-inch  (35.56  cm.)  ball  with  a  100- watt  tungsten  fila- 
ment lamp,  and  Fig.  4  the  same  with  a  250-watt  lamp. 

It  will  be  noted  from  a  study  of  Figs.  2,  3  and  4  that  a  change 
in  the  size  of  lamp  or  of  the  size  of  ball  has  but  little  effect  in 
the  resulting  distribution  curves.  It  will  be  further  noted  that 
the  curves  all  tend  toward  a  circular  distribution,  denoting  good 
diffusion,  but,  at  the  same  time,  very  little  redirection  of  light 
in  useful  zones. 

Fig.  5  is  a  picture  of  a  blown  one-piece  opal  ball  (Class  II), 
the  density  being  the  same,  but  the  thickness  being  less  than  in 
the  pressed  ball  just  considered.  Blown  opal  balls  of  this  den- 
sity and  thickness  are  regularly  furnished  with  the  outside  sand- 
blasted or  roughened  to  increase  diffusion.  Fig.  6  shows  the 
photometric  curve  of  the  12-inch  (30.48  cm.)  size  tested  with  a 
100-watt  lamp.  It  will  be  noted  that  the  curve  is  less  modified 
from  the  bare  lamp  curve  than  in  the  former  case  showing  that 
the  diffusion  is  not  as  good. 

This  is  further  emphasized  by  Fig.  7,  the  right-hand  curve 
being  that  of  the  pressed  ball  and  the  left-hand  one  the  blown 
ball.  The  difference,  while  noticeable,  is  not  striking.  The 
use  of  such  balls  instead  of  the  pressed  type,  would  mean  that 
a  greater  flux  would  strike  the  side  walls  and  less  fall  upon  the 
ceiling.  An  appreciably  greater  flux  is  also  shown  in  the  lower 
hemisphere.  This  type  of  ball  is  used  largely  for  street  lighting 
purposes  and  it  will  be  seen  that  it  has  a  low  absorption,  namely, 
about  14  per  cent,  as  compared  with  an  absorption  of  about  24 
per  cent,  in  the  case  of  the  pressed  type.  A  comparison  of  the 
flux  in  the  different  zones  as  given  under  the  curves  shows  an 
increase  in  the  flux  from  zero  to  6o°  of  approximately  22  per 
cent.,  and  from  zero  to  900  of  19  per  cent.  It  would  seem, 
therefore,  that  in  most  cases,  except  perhaps  from  the  standpoint 
of  appearance,  the  blown  ball,  being  much  thinner,  is  preferable 
for  use  when  there  is  no  objection  to  a  slight  image  of  the  fila- 
ment through  the  glass. 


450  TRANSACTIONS  I.  $.  S.— PART  II 

Fig.  8  shows  a  blown  opal  acorn  shape  (Class  III)  enclosing 
unit  made  of  the  same  glass  as  the  others  so  far  considered,  and 
roughed  outside.  From  its  distribution  curve,  Fig.  9,  it  will 
be  noted  that  due  to  its  shape,  there  is  a  slight  reflecting  power 
which  throws  more  flux  in  the  lower  hemisphere.  The  absorp- 
tion is  somewhat  greater  than  the  blown  ball,  but  less  than  the 
pressed  type,  as  might  be  expected.  Compared  with  the  blown 
ball,  the  acorn  gives  about  the  same  flux  below  the  horizontal 
but  10  per  cent,  more  in  the  0-600  zone. 

Fig.  10  is  a  cased  one-piece  opal  ball  (Class  IV)  and  Fig.  11 
the  curve  of  the  12-inch  (30.48  cm.)  size  tested  with  a  150-watt 
lamp.  Comparison  should  be  made  between  these  and  the  blown 
opal  ball,  a  curve  of  which  is  shown  in  Fig.  6.  It  will  be  noted 
that  the  flux  below  the  horizontal  is  practically  the  same  but 
that  there  is  less  above  the  horizontal,  resulting  in  an  increase  in 
absorption  of  about  5  per  cent.  The  diffusion,  however,  with  the 
cased  ball  is  much  better  than  in  the  blown  opal. 

Fig.  12  is  a  leaded  opal  ball  (Class  V),  the  curve  of  which, 
with  a  150-watt  lamp,  is  shown  in  Fig.  13.  It  will  be  noted  that 
the  absorption  is  practically  the  same  as  that  of  the  14-inch 
(35.56  cm.)  pressed  ball  (Fig.  3  and  4).  The  distribution 
curve  is  somewhat  different,  however,  there  being  slightly  more 
light  on  the  horizontal  and  less  light  directly  above  and  below. 
The  results  of  this  test  are  more  or  less  surprising  inasmuch  as 
there  is  a  general  idea  that  most  diffusing  leaded  glass  absorbs  a 
large  percentage  of  light.  It  would  seem  that  this  particular  type 
of  leaded  glass  has  no  greater  absorption  than  the  ordinary  light 
density  pressed  opal  ball. 

Fig.  14  is  the  curve  of  a  12-inch  (30.48  cm.)  ground  glass  ball 
(Class  VI)  with  100- watt  lamp.  But  little  alteration  in  the 
curve  of  the  bare  lamp  is  made,  although  some  diffusion  is  ob- 
tained. The  diffusion,  however,  is  quite  different  from  that  ob- 
tained with  the  opal  ball,  Fig.  6.  A  comparison  of  the  two 
curves  shows  a  somewhat  lower  absorption  in  the  case  of  the 
ground  glass  ball  than  with  the  opal  blown  ball,  but  a  somewhat 
lower  efficiency  in  the  zero  to  6o°  zone  and  more  light  at  angles 
near  the  horizontal. 


V 


V* 


>A 


Fig.  i. —Two-piece  pressed  opal 
ball  (Class  I). 


1 

/ 
/ 
1 

\             \ 

\           \ 
1          1 

1 
J             / 

Fig.  2.— Photometric  curve  of  io-inch  two- 
piece  pressed  opal  ball  tested  with  ioo- 
watt  lamp. 


Kig.  3.— Photometric  curve  of  14-inch  two- 
piece  pressed  opal  ball  tested  with 
100- watt  lamp. 


/ 

1 

1 

- 

- 

- 

- 

- 

> 

\ 
\ 
\ 

\ 

'                  1             ' 

\ 

\ 
\ 

\ 

- 

- 

- 

- 

' 

< 

/ 
/ 
/ 

1 

V         /       ' 

\              /            / 

Fig.  4. — Photometric  curve  of  14-inch 
two-piece  pressed  opal  ball  tested 
with  250-watt  lamp. 


Zonai,  Flux  with  1,000  Lumen  Lamp. 

Zone 0-60  0-90            90-180  0-180 

Lamp  alone 203  512                4S8  i.oco 

Fig.  2 182  379                385  764 

Fig.  3 190  384                335  7 '9 

Fig.  4 l82  374                346  720 


Fig.  5.— One-piece  blown  opal  ball 
(Class  II). 


x 

/ 

/ 
/ 

\ 
\ 
\ 

/ 
/ 
1 

I 
\ 

l\ 

/ 

/ 

\ 

^~^- 

s^^ 

Fig.  6. — Curve  of  12-inch  one-piece  blown  opal 
ball  tested  with  100-watt  lamp. 


/       "-SBL 


Fig.  8. — Blown  opal  acorn  (Class  III). 


*'          ^-^~T^ 

/^^\ 

X 

/ 

\        1 

1         / 

\        1                      ■ 

\ 

1       1 
1      1 
/     / 

Fig.  9.— Curve  of  12-inch  blown  opal  acorn 
tested  with  150-watt  lamp. 


Fig.  10.—  Cased  opal  ball  (Class  IV). 


Fig.  11. — Curve  of  12-inch  cased  opal  ball 
tested  with  150-watt  lamp. 


D* 


^o 


Fig.  12.— Leaded  opal  ball  (Class  V). 


Fig.  13.— Curve  of  12-inch  leaded  opal  ball 
tested  with  150-watt  lamp. 


Fig.  15.— Prismatic  deep  reflector-bowl 

(Class  VII). 




/                           \ 

!      r^ 

y~\         \ 

\          / 

\        /          , 

\                          / 

V            /                        / 

\                                  >v             / 

\  /            / 

Fig.  16. — Curve  of  14-inch  prismatic  deep  re- 
flector-bowl tested  with  250-watt  lamp. 


Zonal  Flux  with  1,000  Lumen  Lamp. 

Zone 0-60  0-90  90-180  0-180 

Lamp  alone 203  512  488  1,000 

Fig-  6 223  469  395  864 

Fig-  9 243  476  347  822 

Fig-  11 222  456  359  815 

Fig.  13 181  396  321  717 

Fig.  14 206  484  419  903 

Fig.  16 412  589  180  769 


Fig.  7.— Comparison  curves  of  Class  II  Fig.  17.— Comparison  curves  of  Class  VII 

and  Class  I,  Figures  6  and  2.  and  Class  I,  Figures  16  and  3. 


Zonal  Flux  with  1,000  Lumen  Lamp. 


Zone 0-60 

Fig.    7— Class  1,  Fig.    2     182 
Class  2,  Fig.    6     223 


Fig.  17— Class  1,  Fig.    3 

Class  7,  Fig.  16     412 


190 


0-90 

379 
469 

384 
589 


90-180 
3S5 
395 

335 
180 


0-180 
764 
864 

719 
769 


Fig  14— Curve  of  12-inch  ground  glass  ball  (Class  VI)  tested  with  100-watt  lamp. 


\f 


/J 


Fig.  18. — Prismatic  deep  refleetor-bowl       Fig.    19. — Curve  of   14-inch   prismatic  deep 
with  special  deep  bowl.  reflector-bowl   with   special  deep  bowl 

tested  with  250-watt  lamp. 


Fig.  20. — Prismatic  shallow  reflector-bowl        Fig.  21. — Curve  of  12  inch  prismatic  shal- 
(ClassVIII).  low  reflector-bowl  tested    with    100- 

watt  lamp. 


ZoNAt  Fivux  with  1,000  Lumen  Lamp. 

Zone 0-60  0-90            90-180  0-180 

Lamp  alone 203  5 1 2                 488  1 ,000 

Fig.  19 420  5S3                 185  768 

Fig-  21 362  554                 169  723 


Fig.  22. — Prismatic  reflector-ball 
(Class  IX). 


Fig.  23. — Curve  of  12-inch  prismatic  reflector- 
ball  tested  with  150- watt  lamp. 


Fig.  24.— Comparison  curves  of 
Class  IX  and  Class  II,  Figures 
23  and  6. 


/ \ 

1                  / 

1 \ 

u             \                    1 

1               / 
\          / 

\           /                     / 
\        /                     / 
x  /                / 

\                1 

v  v 

Fig.  25. — Curve  of  12-inch  prismatic  reflector- 
ball,  reflector  satin  finished  inside,  tested 
with  150-walt  lamp. 


Zonal  Flux  with  1,000  Lumen  Lamp. 

Zone 0-60              0-90            90-1  So  0-1S0 

Larnpalone 203                 512                 488  1,000 

Fig.  23  (Class  IX)  •■••     4"55                 618                 103  721 

Fig.    6  (Class  II) 223                469                395  864 

Fig.  25    356                552                 1S2  735 


LANSINGH  :     ENCLOSING    GLASSWARE) 


451 


Fig.  26.— Curve  of  14-inch  prismatic  reflector-ball  tested  with  400-watt  lamp 
with  center  of  filament  at  center  of  ball. 


^  V 

Fig.  27.— Curve  of  same  unit  as  Fig.  26  but  with  filament  2  inches  higher. 

Zonal  Flux  with  1,000  Lumen  Lamp. 

Zone 0-60             0-90            90-180  0-180 

Lamp  alone 203                512                488  1,000 

Fig.  26 239                522                '.63  786 

Fig.  27 332                551                 168  719 


With  prismatic  enclosing  units,  a  most  decided  change  in  the 
resulting  distribution  may  be  noted.  With  opal  and  ground 
glass  diffusing  units,  the  tendency  of  all  curves  is  to  become  cir- 
cular in  shape,  with  increasing  diffusion,  but  in  practically  no 
case  is  there  a  large  increase  in  useful  flux  below  the  horizontal. 
In  only  three  cases,  Figs.  6,  9,  and  II,  is  the  6o°  flux 
greater  than  that  of  the  bare  lamp,  being  respectively  223  lumens, 


452  TRANSACTIONS  I.   E.   S. PART  II 

243  and  222,  as  compared  with  203  for  the  lamp  alone.  The 
lower  hemispherical  flux  in  every  case,  however,  is  considerably 
less  than  that  of  the  bare  lamp,  averaging  18  per  cent,  less,  ex- 
clusive of  the  ground  glass  ball. 

The  most  pronounced  redirecting  effect  is  with  the  blown  acorn 
(Fig.  9)  where  24.3  per  cent,  of  the  lamp  flux  is  within  the 
O-600  zone.  A  comparison  of  this  with  Fig.  23  shows  the  re- 
markable difference  which  can  be  obtained  by  the  use  of  specu- 
larly reflecting  media.  In  the  latter  case,  no  less  than  45.5  per 
cent,  of  the  flux  is  within  this  zone,  an  increase  of  77  per  cent. 

Fig.  15  is  a  prismatic  deep  reflector-bowl,  with  a  diffusing 
bottom,  a  photometric  curve  of  which  is  shown  in  Fig.  16.  This 
distribution  is  an  intensive  type  and  by  far  the  greater  flux  is 
below  the  horizontal :  41  per  cent,  is  within  the  6o°  zone,  and 
59  per  cent,  is  below  the  horizontal.  The  total  absorption  is  about 
23  per  cent. 

Fig.  17,  a  comparison  of  Figs.  4  and  16,  is  interesting.  The 
former  curve,  that  of  the  two-piece  pressed  opal  sphere,  shows 
only  18.2  per  cent,  of  the  flux  in  the  6o°  zone  as  against  41.2  per 
cent,  for  the  prismatic  deep  reflector-bowl ;  that  is,  the  latter  unit 
utilizes  about  two  and  a  quarter  times  as  much  flux  in  this  zone. 
The  lower  hemispherical  flux  is  58.9  and  37.4  for  the  reflector- 
bowl  and  opal  ball,  respectively,  with  an  absorption  of  21.3  per 
cent,  and  28  per  cent,  in  each  case. 

Fig.  18  shows  a  modification  of  the  unit  shown  in  Fig.  15,  the 
shallow  bottom  bowl  being  made  much  deeper  in  the  form  of  an 
acorn.  The  resulting  photometric  curve,  Fig.  19,  is  not  greatly 
altered,  although  the  general  shape  looks  more  like  the  extensive 
type  of  distribution  than  the  intensive  type.  The  principal  dif- 
ference is  the  flux  between  zero  and  300. 

Fig.  20  shows  a  shallow  prismatic  reflector-bowl  unit,  the  bowl 
having  shallow  external  redirecting  prisms,  and  Fig.  21  the  re- 
sulting distribution  curve.  It  will  be  noted  that  this  gives  a 
broad  distribution,  much  broader  than  that  shown  in  Fig.  16  and 
19,  but  the  zonal  and  overall  efficiency  is  somewhat  lower.  Of 
the  total  lamp  flux,  36  per  cent,  is  within  the  6o°  zone 
and  55  per  cent,  below  the  horizontal  while  the  absorption  is 
about  28  per  cent. 


LANSINGH  I     ENCLOSING   GLASSWARE  453 

Fig.  22  shows  a  prismatic  reflector-ball,  the  reflector  being 
separate  from  the  blown  globe  and  resting  on  a  shoulder  made 
for  that  purpose.  The  lower  half  of  the  ball  is  satin  finished, 
but  the  part  under  the  reflector  is  clear.  Fig.  23  shows  the  dis- 
tribution curve  obtained  from  this  unit  with  the  lamp  in  the  stand- 
ard position.  It  shows  a  remarkable  control  of  light,  45.5  per 
cent,  of  the  lamp  flux  being  within  the  zero  to  6o°  zone  and  61.8 
per  cent,  below  the  horizontal ;  while  the  total  absorption  is  about 
28  per  cent.  A  comparison  (Fig.  24)  of  this  unit  with  the  blown 
opal  ball  (Class  II,  curve  shown  in  Fig.  6)  is  even  more  striking 
than  the  comparison  in  Fig.  17.  The  flux  in  the  zero  to  6o° 
zone  is  45.5  per  cent,  of  total  lamp  flux  in  the  case  of  the  re- 
flector-ball and  only  22.3  per  cent,  with  the  opal  ball,  the  corres- 
ponding figure  for  the  lower  hemisphere  being  61.8  per  cent, 
and  46.9  per  cent,  while  the  absolute  absorption  is  27.9  per  cent, 
and  13.6  per  cent,  respectively.  The  relatively  low  upper  hem- 
ispherical flux  shown  in  this  reflector-ball  is  quite  remarkable 
and  strikingly  illustrates  the  light  control  which  is  possible  in 
prismatic  combinations,  even  with  a  secondary  globe  between  the 
light  source  and  the  active  reflector  as  is  the  case  in  the  larger 
sizes  of  this  type  of  unit. 

Fig.  25  shows  the  distribution  from  the  same  unit  but  with 
the  reflector  satin  finished  on  the  inside.  It  will  be  noted  at 
once  that  there  is  a  considerable  decrease  in  the  flux  below  the 
horizontal  and  an  increase  in  that  above,  35.6  per  cent,  being 
within  the  zero  to  6o°  zone  and  55.2  per  cent,  below  the  horizon- 
tal. While  this  is  a  considerable  decrease  from  that  obtained 
with  the  clear  reflector,  it  should  be  compared  with  the  best  of 
the  opal  types,  that  is  to  say,  the  acorn  shape  where  only  24.3  per 
cent,  of  the  light,  as  compared  with  35.6  per  cent,  in  this  case, 
is  below  6o°.  Where  the  softer  appearance  of  ground  glass  is 
desired,  a  unit  of  this  type  combines  efficiency  and  such  an  ap- 
pearance. 

Considerable  variation  in  distribution  .may  be  obtained  in  the 
case  of  the  reflector-balls  shown  in  Fig.  22  as  will  be  seen  in 
Figs.  26  and  27  where  a  14-inch  (35.56  cm.)  ball  of  this  type 
with  a  400-watt  lamp  was  used  in  both  cases  but  with  the  center 
of  the  filament  in  the  first  case  at  the  center  of  the  ball  and  in  the 
6 


454  TRANSACTIONS  I.  £.   S. — PART  II 

second  case,  2  inches  (5.08  cm.)  above.  The  difference  in  distri- 
bution is  quite  remarkable  and  shows  that  by  proper  placement 
of  the  lamp,  wide  variations  in  distribution  can  be  obtained. 

SUMMARY. 

From  a  distribution  standpoint,  enclosing  glassware  can  be 
divided  into  two  distinct  classes,  one  purely  transmitting  and  dif- 
fusing, such  as  ground  glass,  opal  glass,  etc.,  and  the  other, 
prismatic  glass,  where  the  principle  of  specular  reflection  is  em- 
ployed. With  glassware  of  the  first  type,  little  except  good 
diffusion  and  low  absorption  can  be  expected,  the  redistribution 
of  light  being  negligible.  In  the  case,  however,  of  prismatic 
glass,  it  is  possible  to  vary  the  distribution  in  accordance  with  the 
wishes  of  the  engineer.  It  is  to  be  noted  further  that  the  ab- 
sorption of  light  in  both  classes  is  about  the  same. 

The  writer  regrets  that  it  has  been  impossible  for  him  to  have 
conducted  depreciation  tests  on  the  different  units,  as  this  phase 
of  the  subject,  namely  the  loss  of  light  and  change  in  distribution 
due  to  dust,  should  not  be  overlooked. 


LANSINGH  :     ENCLOSING    GLASSWARE 


455 


■ 

^                    ^ 

_,  cd             **  X 

TJ 

.                       u,  o      ^3  a 
<:                   d  i_      4)  is  oj 

s 

OJ 

0.0 

C 

Form 

rof  filar 
e  cente 

Center 
ter  of  fi 
ive  cent 

as 

iJ  o 

a. 

f       .    -    . 

"3     -  -  - 

41 

u 

" 

- 

- 

41    >                  =2 

"  o           a-2 

C  .O             (J   « 

0)     Cd                   *"*«. 

O                                N 

D.  & 

5  « 

o      o  o  o 
0       o  >o  o 

O 

in 

O 

8     8 

o       o       o       o           o . 
>o      io      o       io     -    o  - 

3 1 

T          M 

PI                PI                -H                PI                          Tj- 

PI           N    10« 

00 

rO 

ON        PI 

r-»      co        >-        io       loso  oo 

«5 

*d-        ■*  Os  On 

rO 

O         00 

PI          ON         rj-         rO         lONO  O 

ro        H   Tf  N 

d 

M 

>i          pi 

NO          rO         PI          O          ONvO  NO 

ro         io  !>■  ■<d" 

tI- 

•+ 

H           Tf 

NO        00        vO          ON       CC    -    i-i 

V   J*. 

«       <J<  M 

CS 

pa 

O          < 

<d     <     <     <j     <yj 

en 

13 

•    H 

(A 

u 

6 

Q  H 

en 

w<5 

a 

o 

CO 

c    . 

4)    O 

Ocj 

CI 

Z    55 

d* 

»    . 

W  o 

<     < 

Q 

S 

.Q   en  -    -    - 

41    cd 

; 

01 
13 

x'i  1 

3   U 

■§.«=       :       :       :  :   : 

■S3 

a 

0) 

•343 

KJ3 

s 

3 

o  M 

Ih 

o 

0) 

.3.2 

O         rf-      PI 

N 

M 

I/O          PI 

Tf"            "*            PI             PI            -        Tf  - 

Q.s 

p»       io      o 

X 

<* 

o 

00 

PI             Tt            PI             PI                      Tf 

«5 

oo      oo  -    on 

LO 

PI 

^r      _       _       w      _    m  - 

rt         TJ--     NO 

vO 

rO        PI          On       vO         -    CO  " 

tO          fO         PI 

cO 

IO 

M 

00         t^        r-»        Ov              On 

s 
o 

"Alba" 
all 

"Alba" 
all 

"cd 

XI 

< 

orn 

lilite" 

all 

id  opal 

all 

id  glass 

all 

"  Holophane- 

Realite" 

Deep  "  Holo- 

phane-Realite  " 

Holophane 

reflector-bowl 

Holophane 

reflector  ball 

II 

u                r 

o 

U    <V  XI 

cd        o 

5>                        --S 

5 

i4     O 

bid 
£55 

PI            rO  •'tNO 

On 

fO       •* 

vO          On        H          ro        lOvO   r~» 

O 

* 

— 

<n 

4J 

i-J 

3 

■oj 

0 

(0 

cri 

o 

A 

O 

CJ 

u 

a, 

O 

Q 

W 

it 

,_, 

<a 

W 

4) 

o 


o  w 


J3 

w 

od 

ca 

M 
4) 

R 

00 

4) 

0) 

'J 

41 

o 

^C 

00 

^4 

a 
H 

o 

00 

> 

s 

p 

--J 

cd 

3 

cd 

5 

Q.3 

0 

H 

8 

H 

N 

1 

456  TRANSACTIONS  I.  E.   S. — PART  II 

DISCUSSION. 

Mr.  H.  S.  Dunning:  In  abstracting  his  paper  Mr.  Lansingh 
brought  up  the  question  of  the  performance  of  incandescent 
lamps  in  enclosed  globes.  In  the  laboratory  of  the  company 
with  which  I  am  connected,  we  have  been  making  recently  a 
series  of  experiments  in  which  we  found  it  necessary  to  burn 
several  of  the  ioo-watt  size  tungsten  lamps  in  a  small  container. 
The  container  was  lined  with  asbestos  and  there  was  very  little, 
if  any,  ventilation.  The  experiment  which  we  were  making 
depended  in  no  way  on  the  maintenance  of  the  lamps,  but  as  a 
matter  of  interest  we  tested  them  from  time  to  time  and  found 
very  little  change.  I  do  not  believe  that  under  ordinary  condi- 
tions the  life  of  the  tungsten  lamp  is  materially  shortened  by 
using  glassware  which  practically  encloses  the  entire  lamp. 

Mr.  S.  G.  Hibben  :  This  paper,  contains  some  very  excellent 
photometric  data  on  the  particular  types  of  enclosing  units  that 
are  shown  here.  But  I  think  I  am  justified  in  saying  that  it  is 
hardly  a  fair  proposition  to  give  the  distribution  curves  of 
totally  enclosing  units  and  having  selected  that  particular  unit 
which  gives  the  most  light  in  a  downward  direction,  to  then  con- 
clude that  every  time,  or  under  all  conditions,  this  particular 
type  of  enclosing  unit  is  the  only  one  that  is  "efficient"  or  justi- 
fied in  its  use. 

The  conclusion  states  that  with  diffusing  glassware,  little  ex- 
cept good  diffusion  may  be  expected,  the  redistribution  of  light 
being  negligible.  This  statement  will  bear  considerable  quali- 
fication. In  the  first  place  the  positioning  of  the  lamp  in  the 
globe  will  greatly  change  the  results,  and  if  in  these  reported 
tests  the  lamps  had  not  been  placed  so  as  to  give  in  some  cases 
the  maximum  downward  reflection,  and  in  others  to  give  dis- 
tinctly different  distribution,  the  conclusions  could  not  have 
had  such  an  apparently  strong  foundation.  Secondly  there  is  a 
change  in  distribution  with  a  change  in  shape,  even  with  diffus- 
ing glassware.  A  conical  or  parabolic  shaped  reflector  made  of 
white  or  the  so-called  opal  glass,  with  a  diffusing  plate  beneath, 
would  in  fact  give  considerable  redirection  of  light.  The  above 
statement  about  negligible  re-distribution  can  correctly  apply 
only  to  a  spherical  shaped  or  a  very  nearly  spherical   shaped 


ENCLOSING   GLASSWARE  457 

diffusing  globe,  of  the  same  quality  and  finish  of  glass  through- 
out. 

The  user  of  a  diffusing  unit  may  have  in  mind  the  attainment 
of  a  re-direction  of  light,  but  he  also  wishes  low  intrinsic  bril- 
liancy of  the  source,  a  considerable  degree  of  ornamentation,  and 
possibly  a  slight  color  effect.  If  he  wishes  downward  reflection 
primarily,  he  uses,  or  should  use,  an  open  bottom  shade.  And  if 
the  enclosing  globe  meets  all  these  requirements  except  the  first, 
it  may  be  used  efficiently  in  the  broad  sense  of  the  word,  and 
its  use  is  justified. 

I  wish  to  add  that,  inasmuch  as  it  is  the  policy  of 
the  society  to  exclude  from  the  Transactions  the  special  trade 
names  of  various  products,  when  such  trade  names  are  excluded, 
the  proper  general  name  ought  to  be  supplied  in  every  case.  I  do 
not  wish  to  criticise  the  present  paper  on  this  point,  but  I  refer 
here  specifically  to  the  word  "opal'  in  describing  glasses  that  are 
not  opal  in  any  sense  of  the  word.  In  fact,  if  the  two-piece  unit 
of  Fig.  i,  were  actually  made  of  opal  glass,  as  is  stated,  then 
each  radius  of  the  photometric  curve  shown  in  Fig.  2  would  be 
reduced  at  least  one-third. 

In  short,  "opal"  is  not  the  proper  term  to  apply  to  the  ma- 
jority of  the  diffusing  units  that  have  been  discussed  here.  I 
would  like  to  suggest  in  this  case  to  have  the  proper  committee 
take  up  this  matter  of  nomenclature. 

The  words  "opal"  and  "opalescent"  have  been  used  rather 
loosely.  "Opal"  describes  a  particular  class  of  glass,  as  does 
"crystal,"  or  "alba."  There  is  a  sharp  distinction  between  opal 
and  alba  glasses.  The  opal  is  a  very  great  absorber  of  light,  and 
in  some  thickness  or  another  will  always  show  a  yellowish-red 
color  of  transmitted  light.  You  might  use  the  term  "Mazda" 
in  speaking  of  all  metal  filament  lamps,  whatever  the  metal  or 
the  burning  efficiency,  and  this  would  be  no  further  wrong  than 
to  misuse  these  terms  I  speak  of. 

This  is  a  point  I  wish  to  bring  out  very  strongly,  that  the  word 
"opal"  be  limited  to  describing  those  glasses  that  are  really  opal, 
and  that  it  be  not  applied  to  all  white  diffusing  glasses. 

Mr.  V.  R.  Lansingh:    (In  reply)  :    Mr.  Hibben  spoke  of  the 


458  TRANSACTIONS  I.  E.   S. — PART  II 

possibility  of  having  different  shapes  in  diffusing  glassware  to 
give  different  photometric  curves.     It  may  be  possible  to  do  this. 

The  paper  does  not  attempt  to  make  any  comments  whatever 
as  to  the  use  of  the  glassware  which  was  tested,  and  consequently 
I  will  say  nothing  with  regard  to  Mr.  Hibben's  remark  on  that 
subject,  as  it  is  extraneous  to  the  paper  itself. 

As  regards  the  use  of  the  word  "opal,"  I  should  be  very  glad 
indeed  to  have  some  better  word,  but  at  the  present  time  I  know 
of  no  more  definite  classification  unless  we  use  the  trade  names 
themselves. 


richtmyer:    photo-electric  cell  ix  photometry      459 
THE  PHOTO-ELECTRIC  CELL  IN  PHOTOMETRY.* 


BY   F.    K.    RICHTMYER. 
Assistant  Professor  of  Physics,  Cornell  University. 


Synopsis:  On  account  of  some  special  peculiarities,  the  photo-elec- 
tric cell  presents  some  interesting  possibilities  for  photometric  use.  Since 
the  discovery  of  the  so-called  photo-electric  phenomenon  in  1888,  the 
process  of  manufacture  of  the  cells  has  been  so  perfected  that  cells  of 
sodium,  potassium,  or  rubidium,  sensitive  to  light  from  the  visible  spec- 
trum, are  readily  obtainable  on  the  market.  Since  the  intensity  of  the 
current  furnished  by  these  cells  is  strictly  proportional  to  the  intensity 
of  illumination  on  the  sensitive  metal  surface,  the  cells  may  be  used  for 
intensity  measurements  over  a  very  great  range.  For  the  lower  inten- 
sities an  electrometer  must  be  used;  for  higher  intensities  a  sensitive 
galvanometer  is  permissible.  Diagrams  of  connections,  and  some  sug- 
gestions for  the  several  methods  of  using  an  electrometer  are  given.  To 
use  photo-electric  cells  for  photometric  purposes  however,  one  must  be 
perfectly  familiar  with  their  peculiarities.  For  example,  the  wave-length 
sensibility  curve  lies  much  farther  toward  the  violet  than  does  the 
luminosity  curve  for  the  human  eye.  So  that  if  used  for  ordinary  photo- 
metry a  specially  selected  set  of  absorbing  screens  must  be  available.  But 
in  spite  of  some  difficulties  attending  its  use,  the  high  sensibility  and  the 
peculiar  action  of  the  cell  as  a  time  integrator  of  light  intensity  make  it 
particularly  desirable  for  special  photometric  purposes. 

The  photometrist  and  student  of  illumination — especially  when 
working  in  the  research  laboratory — frequently  meets  conditions 
which  make  it  desirable  to  use  an  apparatus  for  measurement 
which  eliminates,  in  part  at  least,  some  of  those  ever-present  diffi- 
culties incidental  to  photometry  by  the  human  eye.  It  may  be 
desired  to  avoid  various  physiological  and  psychological  effects. 
Precision  greater  than  that  available  by  eye  measurement  may  be 
necessary.  The  illumination  may  be  of  too  short  duration  for 
observation  by  ordinary  means.  Of  the  various  devices  avail- 
able for  use  in  such  cases  none  deserves  greater  attention  than 
the  so-called  photo-electric  cell.  In  the  hands  of  one  who  fully 
understands  its  use  and  its  limitations,  it  should  prove  a  most 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineeting 
Society.  Pittsburgh,  Pa..  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


460  TRANSACTIONS  I.  £.   S. — PART  II 

valuable  addition  to  any  laboratory.  It  is  the  purpose  of  this 
paper  to  point  out,  in  very  brief  outline,  some  of  its  character- 
istics, peculiarities,  and  advantages,  and  to  give  several  specific 
examples  for  its  use. 

HISTORY. 
Those  who  have  not  followed  the  development  of  the  photo- 
electric cell  may  be  interested  in  knowing  that  as  far  back  as 
1888,  Hertz — to  whom  perhaps  more  than  any  other  we  owe  the 
wireless  telegraph — discovered  that  under  certain  conditions  a 
metal  plate,  connected  to  the  negative  terminal  of  an  electric 
generator,  the  positive  terminal  of  which  was  grounded,  and 
illuminated  by  the  ultra-violet  light  from  a  spark,  would  discharge 
into  the  air  or  to  a  nearby  grounded  wire  a  continuous  stream  or 
current  of  negative  electricity.  In  the  early  '90's  this  interesting 
relation  between  light  and  electricity  was  the  subject  of  much 
investigation.  Among  other  things,  it  was  shown  by  Elster  and 
Geitel,1  that  the  alkali  metals,  particularly  sodium  and  potassium, 
were  sensitive  to  light  from  the  visible  spectrum,  especially  to 
blue  and  violet.  On  account  of  the  rapidity  with  which  these 
metals  oxidize  when  exposed  to  the  air,  it  was  necessary  to  study 
them  in  an  atmosphere  of  some  inert  gas.  For  this  purpose, 
Elster  and  Geitel  devised  a  method,  which  has  since  been  per- 
fected, of  pouring  the  metals,  in  molten  form,  into  a  small  glass 
bulb  containing  hydrogen  or  helium  and  having  the  necessary 
external  electrical  connections.  These  are  the  photo-electric 
cells  which  may  now  be  obtained  on  the  market  in  a  great  variety 
of  forms  for  a  very  reasonable  price. 

METHODS  OF  USE. 
Fig.  i  represents  diagrammatically  the  principles  involved  in 
using  the  photo-electric  cell.  A  battery  B  has  its  positive  ter- 
minal grounded,  and  its  negative  terminal  connected  to  the  alkali 
metal  S  contained  in  the  cell  C.  A  wire  is  sealed  in  through  the 
glass  and  connected  through  a  sensitive  galvanometer  or  other 
measuring  instrument  G  to  earth.  If  a  beam  of  light  be  now 
allowed  to  fall  on  the  metal  surface  in  the  direction  of  the  arrow 
a  current  of  electricity  will  flow  through  the  galvanometer. 

1  See  Annalen  der  Physik,  Vol.  43.  p.  225  (1891). 


richtmyer:    photo-electric  cell  in  photometry      461 

The  strength  of  this  current  will  depend  on  several  things: 
It  increases  as  the  electromotive  force  of  the  battery  B  is  in- 
creased up  to  a  certain  point — at  least  this  is  true  for  most  cells 
obtainable  on  the  market."  It  depends  very  greatly  on  the  color 
(wave-length)  of  the  incident  light.  But  if  these  two,  i.  e., 
electromotive  force  and  color,  are  maintained  constant  the 
strength  of  the  current  is  absolutely  proportional  to  the  intensity 
of  illumination  on  the  metal  surface  over  an  enormous  range  of 
intensities.3 


4hHHH^ 


ck 


# 


Fig.  1.— Diagram  showing  the  essential  connections  for  using  the 
photo-electric  cell.  B  is  a  battery,  its  negative  terminal  con- 
nected to  the  sensitive  metal  S.  The  receiving  wire  or  elec- 
trode C  is  connected  through  a  sensitive  galvanometer  to  earth. 


The  currents  furnished  are  in  most  cases  comparatively  small. 
Only  the  higher  illuminations  (say  20  or  30  foot-candles  for 
white  light)  produce  currents  large  enough  to  be  measured  by  a 
sensitive  galvanometer,  the  connections  then  being  essentially  as 
shown  in  Fig.  1.  These  currents  are  of  the  order  of  magnitude 
of  io-9  amperes.  For  the  more  common  illuminations  (say  from 
G.0005  foot-candles  up)  a  sensitive  electrometer  is  necessary. 
Although  the  latter  instrument  is  somewhat  more  troublesome  to 
handle  than  a  galvanometer,  the  possibilities  of  this  method  of 
measurement  more  than  justify  its  use.4 

One  may  use  an  electrometer  in  either  of  two  ways.  If  the 
illuminations  are  very  small  the  electrometer  should  be  connected 

2  When  using  E.  M.  F.'s  larger  than  a  few  volts  the'  current  obtained  is  due  in  part  to 
the  ionization  in  the  gas. 

3  See  paper  by  the  writer,  Phystcal  Review.  Vol.  29,  p.  404  (1909). 

4  No  attempt  will  be  made  here  to  discuss  the  difficulties  incident  to  the  use  of  an 
electrometer.  They  are  by  no  means  insurmountable,  and  those  interested  are  referred 
to  the  various  treatises  on  the  instrument. 


462 


TRANSACTIONS  I.  t.   S. — PART  II 


directly  in  the  circuit  as  shown  in  Fig.  2.  Here  the  electrometer 
Q  replaces  the  galvanometer  of  the  previous  connection,  and  a 
key  k  makes  it  possible  to  discharge  or  insulate  the  pair  of  quad- 
rants to  which  it  is  connected.  A  condenser  M  may  be  used 
in  parallel  with  the  electrometer  to  vary  its  current  sensibility 
as  required.  One  of  the  several  types  of  air  condensers  is  most 
successfully  used. 

With  this  set-up  the  procedure  is  essentially  as  follows :  The 
illumination  to  be  measured  is  allowed  to  fall  on  the  cell  and  the 
key  k  is  kept  closed  until  stationary  conditions  have  been  reached. 
k  is  then  opened  and  the  rate  of  drift  of  the  electrometer  meas- 


Fig  2. — Showing  the  method  of  connecting  the  photo- 
electric cell  when  using  an  electrometer  Q  and  a 
condenser  M  in  place  of  the  galvanometer.  A  key  k 
makes  it  possible  to  insulate  or  ground  the  elec- 
trometer. 


ured  by  a  stop  watch  or  chronograph.  If  C  is  the  capacity  of 
the  electrometer  system,  K  its  constant  (i.  e.,  the  number  of  volts 
necessary  to  deflect  it  one  scale  division)  and  R  the  measured 
rate  of  drift  in  divisions  per  second,  then  the  current  I  is  given  by 

I  =  C-K-R 
The  capacity  used  with  the  electrometer  should  be  such  that  the 
rate  of  drift  does  not  exceed  two  or  three  millimeters  per  second. 
However,  with  a  good  telescope  and  scale  and  proper  insulation 
and  screening  for  the  electrometer  rates  of  drift  as  low  as  a  few 
hundredths  of  a  millimeter  per  second  can  be  accurately 
measured. 

This  "rate  of  drift"  method  has  its  objections,  partly  due  to 
the  fact  that  a  comparatively  long  time  is  required  for  each  obser- 


RICHTMYER  :     PHOTO-ELECTRIC  CEEL  IN  PHOTOMETRY        463 

vation.  The  writer  has  used,  as  a  substitute  for  it  in  some  cases, 
the  "ballistic"  method.  This  requires  in  front  of  the  cell  a 
shutter  which  can  be  opened,  mechanically  or  otherwise,  for  a 
definite  time  interval.  In  this  method,  having  previously  opened 
k,  the  shutter  in  front  of  the  cell  is  opened,  allowing  the  illumina- 
tion to  fall  on  the  cell  for  the  desired  time  interval,  which  must 
be  accurately  measured.  During  this  interval  the  electrometer 
has  been  given  a  certain  charge,  in  consequence  of  which  a  per- 
manent deflection  is  observed  after  the  shutter  is  closed.  If  S 
is  this  deflection,  T  the  time  of  exposure,  and  C  and  K  have  the 
same  values  as  before  the  current  is  given  by 

I  =  C-KS/T 

This  ballistic  method  is  so  useful  for  such  a  great  variety  of 
laboratory  measurements  that  a  special  discussion  of  some  of  its 
features  seems  desirable.  Experiment  seems  to  show  that  at 
least  for  moderate  illuminations  the  quantity  of  electricity  dis- 
charged by  such  a  cell,  subjected  to  a  constant  illumination,  is 
accurately  proportional  to  the  time  of  exposure,  even  when  the 
duration  of  the  illumination  is  only  a  small  fraction  of  a  second. 
This  makes  it  possible  to  make  "snapshot"  measurements  of  the 
candle-power  of  a  fluctuating  light  source.  The  magnitude  of 
the  quantities  concerned  are  indicated  by  an  experiment  by  the 
writer  in  which  a  deflection  of  several  centimeters  was  obtained 
by  an  exposure  to  a  Nernst  glower  of  0.001  second.  The  glower 
was  two  or  three  feet  from  the  cell.  It  is  unnecessary  to  empha- 
size further  the  possibilities  in  this  direction,  for  there  are 
numerous  experiments  where  instantaneous  candle-powers  are 
desirable  under  conditions  where  measurement  by  eye  is  almost 
impossible.  Also,  one  may  easily  reverse  the  process  and  measure 
a  time  interval,  as  for  example  the  speed  of  a  photographic 
shutter. 

When  used  with  this  ballistic  method  the  photo-electric  cell 
acts  as  a  time  integrator  of  light  intensity,  in  much  the  same  way 
as  the  electrolytic  cell  acts  as  a  time  integrator  for  the  electric 
current.  In  other  words,  suppose  it  is  desired  to  find  the  average 
intensity  of  a  fluctuating  light  source  over  a  certain  time  interval, 
long  or  short.     One  would  simply  have  to  expose  the  cell  to  the 


464  TRANSACTIONS  I.  £.   S. — PART  II 

light  source  for  the  desired  interval  and  read  the  resulting  deflec- 
tion. This  divided  by  the  time  gives  the  average  rate  of  deflec- 
tion. By  comparing  this  rate  with  that  produced  by  a  known 
light  source  the  average  candle-power  of  the  first  source  is  at 
once  obtained. 

A  second  method  of  using  the  electrometer  in  connection  with 
the  photo-electric  cell  is  to  connect  the  instrument  to  the  ter- 
minals of  a  high  resistance,  through  which  the  current  flows  on 
its  way  to  earth.  The  principle  of  this  method  is  of  course 
identical  with  the  method  of  measuring  current  by  a  voltmeter 
and  resistance.  Using  this  device  Nichols  and  Merritt5  have  with 
great  rapidity  measured  the  densities  of  a  series  of  photographic 
negatives.  While  this  method  has  the  advantage  of  giving  the 
current  directly  by  the  steady  deflection  of  the  electrometer,  its 
use  is  limited  to  cases  where  fairly  high  intensities  are  available. 

SENSIBILITY  WAVE-LENGTH  CURVE. 

The  manner  in  which  the  photo-electric  current  depends  on 
the  wave-length  or  wave-length  composition  of  the  incident 
light  is  a  crucial  question  when  we  are  considering  the  appli- 
cation of  the  photo-electric  cell  to  photometry.  Numer- 
ous investigators  seem  to  agree  that  as  ordinarily  used  the  cell 
obtainable  on  the  market  has  a  wave-length  sensibility  curve 
which,  while  agreeing  fairly  well  in  shape  with  the  luminosity 
curve  of  the  human  eye,  is  nevertheless  quite  different  in  position, 
having  its  maximum  much  farther  toward  the  violet  end  of  the 
spectrum.  Fig.  3,  I  and  II,  shows  the  sensibility  curve  as  deter- 
mined some  time  ago6  by  the  writer  for  one  of  the  cells  then 
obtainable.  I  is  the  curve  as  observed  without  making  any  cor- 
rections, and  II  is  the  result  of  making  the  correction  for  the 
energy  distribution  in  the  spectrum  of  the  source  used  (curve 
III).  It  is  seen  that  the  maximum  of  curve  II  occurs  at  about 
0.46  [i,  although  its  exact  position  is  left  in  some  doubt  on  account 
of  the  difficulty  of  obtaining  a  satisfactory  curve  for  the  energy 
distribution  of  acetylene,  the  source  used,  in  the  blue  and  violet. 
Later  experiments  by  other  investigators  seem  to  show  that  the 

6  Physical  Review,  Vol.  34,  p.  476  (i912)- 
6  Physical  Review,  Vol.  30,  p.  3S5  (1910). 


richtmyer:    photo-electric  cell  in  photometry      465 

maximum  is  even  farther  toward  shorter  wave-lengths  than  here 
indicated. 

This  high  sensibility  to  blue  and  violet  constitutes  the  chief 
objection  to  the  general  use  of  the  photo-electric  cell  for  all 
photometric  measurements.  One  can  readily  see  that  a  slight 
preponderance  of  blue  in  the  light  as  measured  by  eye,  would 
make  a  vast  difference  in  the  measurement  by  the  photo-electric 
cell,  unless  the  latter  measurement  were  made  by  use  of  a  very 
carefully  selected  set  of  absorbing  screens,  so  chosen  as  to  give 


20 


I5v 


10  = 


in 

!    1 

\ 

2.0 

n-ji 

I- 

m 

"t     ' 

■  Y* 

|j*S 

f\ 

I   \  I 

J  1 

i  \   \l 

\ 

f 

1  \   ii 

-i.5 

/  ■  A 

l\  1  t\ 

\  7  * 

V 

to 

1 

i 

1 

i 

l\ 

\  1 

1 

m 

1  / 

,    1 

'  1 1 

1 7 

1  / 

fr 

I    1 

hh- 

0.4  y 


0.5p  .0.6  p 

WAVE    LENGTH 


0.7y 


Fig.  3. — Showing  the  wave-length  sensibility  curve  for  a  sodium 
cell.  Curve  I  is  the  directly  observed  data.  Curve  II  is  the 
corrected  curve,  correction  for  the  variable  dispersion  of  the 
prism  used  and  for  the  variable  energy  in  the  spectrum  of 
acetylene  having  been  made.  Curve  III  shows  the  energy  dis- 
tribution in  acetylene  (after  Nichols  and  Merritt). 


the  cell  a  sensibility  curve  approximating  that  of  the  human  eye. 
On  account  of  the  high  absorbing  power  which  such  a  set  of 
screens  must  necessarily  have,  the  sensibility  of  the  device  would 
be  greatly  reduced. 

For  this  reason  it  seems  probable  that  the  most  useful  applica- 
tions of  the  photo-electric  cell  at  the  present  time  must  be  limited 
either  to  cases  involving  monochromatic  or  isochromatic  light 
sources,  or  else  to  investigations  where  the  actual  intensity,  pho- 


466  TRANSACTIONS  I.  E.   S. — PART  II 

tometrically  measured,  is  of  only  secondary  importance.  If,  for 
example,  one  were  making  a  series  of  "stroboscopic"  snap-shots 
of  an  alternating  current  arc  over  one  cycle,  he  would  not  expect 
the  resulting  curve  to  agree  either  in  magnitude  or  position  of  its 
maximum,  with  the  curve  obtained  by  ordinary  means.  It  is 
possible,  however,  to  isolate  any  wave-length  from  such  an  arc 
by  means  of  a  spectrometer,  and  by  means  of  the  photo-electric 
cell  follow  it  during  one  cycle  under  conditions  of  intensity  and 
color  which  would  render  eye  measurement  impossible.  Further- 
more, it  is  to  be  remembered  that  in  the  blue  and  violet  region, 
where  eye  measurement  becomes  very  difficult,  the  cell  becomes 
xnost  sensitive. 

SOME  DIFFICULTIES. 

In  addition  to  the  difficulties  previously  mentioned,  namely, 
the  great  care  which  must  be  exercised  in  studying  differently 
colored  illuminations  and  also  the  fact  that  the  electrometer  is 
somewhat  more  troublesome  to  handle  than  a  galvanometer,  there 
exists  one  great  source  of  trouble  due  apparently  to  the  fact  that 
a  photo-electric  cell,  even  in  the  dark,  will  allow  a  small  current 
to  flow  if  an  electromotive  force  be  impressed  on  its  terminals. 
This  seems  to  be  caused  by  the  walls  of  the  cell  possessing  a 
resistance  which  is  of  course  not  infinite.  This  leakage  current 
is  fortunately  too  small  to  be  of  serious  disturbance  when  one  is 
measuring  intensities  of  several  foot-candles.  But  when  working 
with  lower  illuminations  of  a  few  hundredths  of  a  foot-candle  the 
leakage  current  becomes  a  serious  source  of  error.  In  fact,  it 
may  become  much  greater  than  the  photo-electric  current  itself. 
In  any  event  either  it  must  be  determined  and  a  corresponding 
correction  made,  or  it  may  be  eliminated  by  charging  the  cell  to 
such  a  potential  that  the  difference  of  potential  between  the  alkali 
surface  and  the  receiving  wire  is  zero.7  in  which  case  there  is  no 
leakage  current.  Later  cells  of  improved  construction  have 
reduced  this  difficulty  very  greatly. 

Although  the  proportionality  between  photo-electric  current 
and  intensity  of  illumination  seems  to  hold  over  a  very  great 
range  of  intensities,  it  is  probably  not  safe  to  expose  any  cell, 

7  See  paper  by  the  writer,  Physical  Review,  Vol.  29.  p.  71  (1909). 


richtmyer:    photo-electric  cell  in  photometry      467 

for  any  length  of  time  at  least,  to  the  very  high  illuminations 
(several  hundred  foot-candles).  A  cell  used  by  the  writer,  gave 
consistent  and  reproducible  results  for  months  when  used  only 
on  moderate  illuminations,  until  in  one  experiment  it  was  exposed 
at  intervals,  totaling  perhaps  30  minutes,  to  a  2,000  candle-power 
arc  two  or  three  feet  away.  For  weeks  afterward  this  cell  gave 
very  erratic  readings  when  used  again  on  lower  illuminations, 
the  currents  being  not  only  not  reproducible,  but  even  varying 
with  a  constant  illumination,  as  if  to  indicate  that  some  sort  of 
instability  had  been  produced  by  the  very  high  intensity. 

In  spite  of  some  serious  difficulties,  the  photo-electric  cell  is 
being  used  more  and  more  as  a  photometric  device  for  special 
purposes.  And  the  results  obtained  certainly  justify  any  added 
effort  necessary  for  its  use. 

DISCUSSION. 

Mr.  J.  L.  Minick:  I  should  like  to  inquire  as  to  the  per- 
manency of  the  photo-electric  cell.  If  a  set  of  readings  were 
taken  as  Dr.  Ives  has  described,  and  the  experiment  repeated  a 
year  later,  would  the  photo-electric  cell  show  any  deterioration 
in  that  time? 

Mr.  S.  L.  E.  Rose:  I  am  extremely  interested  in  this  paper, 
and  there  are  one  or  two  questions  that  I  would  like  to  ask; 
I  would  like  to  know  if  the  cell  is  sufficiently  developed  to  be 
tried  out  commercially.  The  laboratory  with  which  I  am  con- 
nected is  ready  to  help  in  its  application  to  commercial  photo- 
metry as  soon  as  the  cell  has  demonstrated  its  practicability. 

Dr.  H.  E.  Ives  :  I  think  to  Professor  Richtmyer  is  due  the 
credit  of  being  one  of  the  first  to  appreciate  the  really  enor- 
mous possibilities  of  the  photo-electric  cell.  He  had  papers  in 
the  Physical  Review  I  think  five  or  six  years  ago  showing  the 
laboratory  applications  of  photo-electric  cells.  I  must  say  that 
the  photo-electric  cell  appeals  to  me  as  the  most  interesting  sub- 
ject in  photometry.  I  have  been  working  on  it  pretty  continuously 
for  about  a  year  and  I  feel  as  a  result  of  that  work,  the  only 
criticism  I  would  make  of  Professor  Richtmyer  is  that  he  is 
not  enthusiastic  enough  about  the  possibilities,  and  perhaps  he 
is  not  sufficiently  impressed  with  the  difficulties  of  the  cell. 


468  TRANSACTIONS  I.  E.   S. — PART  II 

I  want  to  say  a  word  about  the  sensibility  of  the  photo-electric 
cell.  Last  night  Dr.  Brashear  told  us  about  the  eye  being  able 
to  look  at  the  sun  for  half  a  second  and  being  able  to  see  a 
brightness  some  four  quadrillions  less.  Up  to  the  present  time 
we  have  had  no  instruments  possessing  anything  like  that 
range.  But  in  recent  work  by  Elster  and  Geitel  the  photo- 
electric current  was  measured  from  illuminations  nearly  that 
from  the  sun  down  to  that  from  a  pin  point  before  a  gas  flame 
20  or  30  feet  away.  They  claim  that  a  direct  proportionality 
holds  between  illumination  and  current  over  the  whole  of  this 
range.  We  have  something  here,  therefore,  that  is  going  to  press 
the  eye  pretty  closely  for  range  and  sensibility. 

Now,  as  to  the  accuracy  and  sensibility  of  the  photo-electric 
cell, — I  have  made  measurement  after  measurement  with  the 
range  not  exceeding  2/ioths  of  one  per  cent.,  the  reading  being 
just  as  easy  as  with  a  first-rate  voltmeter.  That  certainly  ap- 
peals I  think  to  people  who  have  had  to  do  much  photometric 
work. 

A  great  deal  of  work  has  been  done  since  Professor  Richt- 
myer's  study  upon  the  theory  of  the  photo-electric  cell.  It  has 
been  found  that  the  maximum  sensibility  lies  in  different  parts 
of  the  spectrum  for  different  metals ;  with  calcium  it  should  be 
just  where  the  eye  is  most  sensitive.  We  have  a  possibility  here 
in  that  cells  can  perhaps  be  made  with  the  color  sensibility  of  the 
human  eye.  We  might  then  use  the  photo-electric  cell  for 
colored  light  photometry  instead  of  having  to  call  on  a  large 
number  of  observers  or  adopt  some  other  roundabout  method. 

There  is  a  great  deal  yet  to  be  done  with  the  cell,  however. 
The  question  of  permanency  has  been  raised  and  will  take  time 
to  settle. 

Now  in  regard  to  the  suggestion  of  putting  the  cell  in 
front  of  a  moving  carriage  and  running  it  around.  If  the 
speaker  would  come  down  to  the  basement  of  the  laboratory 
I  am  connected  with  and  watch  what  happens  to  the  electrometer 
when  somebody  sneezes  in  the  room  above,  he  would  realize  that 
there  is  yet  some  work  to  be  done  before  we  can  carry  a  cell 
around  on  a  moving  carriage.  Nevertheless,  the  work  I  am  do- 
ing   with    a    spectro-photometer    demands    extreme    sensibility. 


PHOTO-ELECTRIC   CELE  IN   PHOTOMETRY  469 

Where  the  light  is  taken  directly,  without  interposing  light  ab- 
sorbing instruments  it  has  been  found  possible  to  use  a  portable 
galvanometer,  so  perhaps  we  are  not  so  far  from  the  suggestion 
as  might  appear. 

As  to  the  cells  being  on  the  market,  they  are  on  the  market; 
but  I  would  not  advise  anybody  to  buy  them  as  they  are  now 
being  made. 

In  summarizing  this  whole  question,  I  want  to  be  very  en- 
thusiastic. I  really  believe  that  for  a  great  deal  of  laboratory 
photometry,  especially  where  lamps  of  different  colors  are  to  be 
compared,  in  a  few  years  we  will  actually  be  using  the  photo- 
electric cell  in  place  of  the  eye,  this  provided  of  course  that  the 
simple  relationship  holds  between  illumination  and  current,  and 
that  the  outstanding  question  of  permanency  and  uniformity  are 
satisfactorily  settled. 


470  TRANSACTIONS  I.  E.  S. — PART  II 

FACTORY  LIGHTING.* 


BY  M.   H.  FX3XNER  AND  A.  O.  DICKER. 


Synopsis:  Ever  since  the  realization  of  the  good  and  bad  effects  of 
illumination,  there  has  always  been  a  great  field  in  factories  for  better 
lighting  conditions.  Better  light  is  as  necessary  as  any  sanitary  require- 
ments and  with  these  it  should  rank  among  the  first.  Foreign  countries 
have  taken  better  illumination  a  little  more  seriously  than  America  has. 
They  have  had  committees  appointed  by  the  government,  whose  duties  are 
to  study  the  effects  of  good  and  bad  light  upon  the  general  health  and 
report  upon  methods  of  bettering  conditions.  Although  the  importance 
of  good  lighting  is  generally  understood,  managers  of  factories  are  never 
willing  to  make  any  decided  changes  from  present  operating  conditions. 
No  matter  how  forcible  the  arguments,  the  first  cost  of  the  installation 
of  a  lighting  system  seems  to  retard  any  change  for  better  illumination. 
We  desire  to  show  how  easily  and  cheaply  conditions  can  be  bettered.  We 
believe  that  it  is  just  a  matter  of  a  short  time  until  the  factory  manager 
will  understand  the  great  importance  of  good  lighting,  and  when  he  does 
he  will  not  be  satisfied  until  he  has  a  lighting  system  that  is  up  to  the 
minute. 

The  aim  of  this  paper  is  to  bring  out  a  few  of  the  most 
important  factors  entering  into  the  design  or  re-design  of  a  light- 
ing system  for  the  factory.  It  is  somewhat  discouraging  to  the 
illuminating  engineer  to  read  article  after  article  dealing  with 
the  methods  used  to  raise  the  sanitary  condition  of  the  factory 
and  when  all  has  been  read  he  asks  himself  "What  about  the 
lighting?"  Ventilation,  cleanliness,  devices  for  safe  operation 
of  machines,  rest  rooms  for  employees  are  all  discussed,  but 
little  or  no  attention  is  given  the  lighting.  It  is  the  hope  of  the 
authors  that  this  paper  will  emphasize  the  fact  that  factory 
lighting  is  a  subject  dealing  directly  with  sanitation  and  that  it 
should  be  considered  as  such.  Why  is  the  lighting  important, 
and  whom  does  it  affect?  Does  it  mean  a  benefit  for  the  central 
station  only,  or  is  it  of  equal  benefit  and  importance  to  employer 
and  employee?  It  seems  just  as  reasonable  to  ask  why  should  a 
factory  be  ventilated  or  why  should  it  ever  be  cleaned  up.     The 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


FLEXNER  AND  DICKER:     FACTORY  LIGHTING  471 

owner  or  manager  would  immediately  say :  "If  I  do  not  ventilate 
the  work  rooms  the  operators  will  become  dull  and  lose  interest 
in  their  work."  Regarding  his  lighting  conditions  he  knows 
naught  and  his  answer  to  a  question  relative  to  his  lighting  con- 
dition, would  very  likely  show  that  he  never  gave  it  much  thought. 
This  is  just  the  man  who  needs  some  information  regarding 
lighting.  He  does  not  realize  that  just  as  many  of  the  headaches 
are  caused  by  poor  lighting  in  factories  as  there  are  from  poor 
ventilation.  This  is  not  intended  to  belittle  the  importance  of 
good  ventilation,  but  is  only  mentioned  to  emphasize  the  fact  that 
general  improvement  of  condition  does  not  end  when  a  factory 
has  been  properly  ventilated  or  properly  cleaned.  It  does  not  end 
until  the  lighting  as  well  as  these  have  been  considered.  One  is 
just  as  important  as  the  other,  since  injury  to  the  eye  from  poor 
lighting  causes  suffering  equal  to  or  even  greater  than  the  sick- 
ness caused  from  poor  ventilation.  In  considering  such  vital 
subjects  this  country  seems  to  be  far  in  the  rear  of  countries  on 
the  other  side.  We're  behind  the  times,  so  to  speak,  and  have 
not  kept  pace  with  France,  England  and  other  European  coun- 
tries, who  are  protecting  their  workmen,  along  these  lines. 

In  1912  the  French  Government  appointed  a  Committee  on 
Hygienic  Aspects  of  Illumination,  composed  of  prominent  physi- 
ologists, oculists,  engineers,  physicists,  and  inspectors  of  fac- 
tories.   The  main  objects  of  this  committee  are: 

(a)  To  study,  from  the  standpoint  of  general  health  and  its 
effects  on  vision,  the  various  methods  of  artificial  lighting  now 
used. 

(b)  To  determine  the  composition  and  quality,  from  a  hy- 
gienic standpoint,  of  the  different  combustible  illuminants,  and 
to  examine  the  effect  of  prejudicial  gases  and  the  amount  of 
heat  developed  thereby, 

(c)  To  fix  a  certain  amount  of  artificial  illumination  to  the 
normal  requirements  of  vision. 

(d)  To  study  the  most  practical  methods  of  measuring  illum- 
ination. 

(e)  To  formulate  recommendations  governing  the  best  means 
of  applying  customary  methods  of  lighting  to  the  chief  varieties 
of  industrial  operations. 


472  TRANSACTIONS  I.  E.   S. PART  II 

(f)  To  present  to  the  Ministry  a  report  on  the  subject  of  short 
sight  and  impairment  of  vision,  and  on  the  best  methods  of 
guarding  against  the  cause  of  myopia. 

It  is  the  result  of  the  investigations  of  such  committees  that 
awaken  the  mind  of  the  manufacturer  to  the  necessity  of  pro- 
viding good  lighting. 

The  first  question  that  might  be  asked  is :  What  is  good  illumi- 
nation, or  what  is  practical  illumination?  Can  we  spot  a  unit 
or  cluster  here  or  there,  put  a  drop  light  over  the  working 
places  in  a  slip-shod  sort  of  manner  and  expect  to  be  satisfied 
with  the  results;  or,  is  it  a  matter  of  knowing  what  to  expect 
from  each  means  of  illumination  and  its  corresponding  reflector 
and  to  fit  in  these  units  to  meet  the  conditions  in  the  factory? 
Our  common  sense  dictates  that  it  is  the  latter.  Our  experience 
teaches  us  that  the  problems  involved  are  often  difficult  of  solu- 
tion and  that  we  must  have  definite  ideas  about  correct  illumina- 
tion before  we  attempt  to  accomplish  satisfactory  results. 

One  authority  defines  good  lighting  as  any  system  which  does 
not  attract  attention  to  the  means  of  illumination,  or  cause  one 
to  wonder  how  the  illumination  was  obtained.  An  analysis  of 
this  yields  the  following  requirements  for  good  lighting: 

First,  that  sources  of  high  intensity  must  not  be  in  the  field 
or  ordinary  vision ;  second,  that  the  amount  of  light  be  sufficient 
for  the  work  to  be  done;  third,  that  the  distribution  of  light 
be  uniform  or  as  nearly  so  as  possible,  and  fourth,  that  the  color 
be  pleasing  to  the  eye.  By  adhering  to  these  principles,  we  will 
not  go  far  wrong  in  laying  out  lighting  installations,  whether 
for  factory  or  for  home,  being  assured  of  good  illumination. 

The  value  of  good  illumination  should  not  be  under-estimated. 
Some  are  contented  to  travel  along  in  the  old  time  worn  ruts  and 
to  leave  well  enough  alone.  Many  believe  that  as  long  as  there 
is  light,  whether  good  or  bad,  the  question  of  lighting  is  settled, 
and  that  the  results  obtained  are  as  good  as  any  light  could 
produce.  This  is  the  wrong  idea,  but  nevertheless  it  is  enter- 
tained by  many  managers  and  officials  of  factories  under  whose 
jurisdiction  the  question  of-  lighting  comes ;  however,  they  must 
realize  sooner  or  later  the  value  of  better  operating  conditions, 
produced  by  good  lighting.    To  do  work,  light  is  necessary ;  with 


ELEXNER  AND  DICKER:     FACTORY  LIGHTING  473 

a  little  light,  a  little  work  can  be  done  and  with  more  light  more 
work  can  be  accomplished.  This  is  very  evident,  and  it  is  easily 
seen  that  no  matter  how  a  shop  is  lighted,  if  it  can  be  better 
lighted,  better  or  more  work  must  result,  up  to  a  definite  per 
cent,  increase  in  efficiency  of  the  workman.* 

What  if  our  Mr.  Official  had  to  go  home  to  a  dimly  lighted 
dining  room?  How  would  he  like  to  read  a  paper  which  neces- 
sitated straining  his  eyes,  or  shave  in  little  or  no  light,  with  his 
face  very  near  the  mirror  and  his  eyes  fixed  in  a  staring  position. 
It  would  not  be  very  comfortable  and  he  could  hardly  give  him- 
self much  of  a  shave;  yet  under  these  conditions  he  expects  his 
men  to  work,  to  turn  out  good  work,  and  make  his  factory  an 
efficient  one. 

There  are  such  things  as  good  and  bad  lighting  installations, 
and  to  the  progressive  official  the  best  should  not  be  too  good  for 
his  men.  However,  the  initial  cost  is  given  first  consideration  and 
is  the  one  stone  that  lies  in  the  path  of  all  changes,  and  therefore 
we  can  but  sum  up  the  reasons  why  it  is  worth  every  cent  that 
is  asked  in  making  a  lighting  installation  a  good  and  efficient 
one. 

Statistics  have  shown  that,  as  the  result  of  better  illumination 
and  a  decreased  strain  on  the  eye,  the  physical  condition  of  the 
workmen  is  better,  they  are  better  satisfied,  imperfections  in  the 
work  have  been  materially  decreased  and  the  factory  output 
increased  from  8  to  15  per  cent. 

Not  only  is  the  general  physical  condition  of  the  workman 
improved  by  better  lighting  but  his  liability  to  accident  is  greatly 
decreased.  Recently  published  statistics  show  that  during  those 
months  of  the  year  in  which  artificial  lighting  must  be  used,  there 
occurs  a  greater  number  of  accidents  than  in  the  light  months. 
The  saving  made  by  good  lighting  in  this  line  alone  will  often 
more  than  repay  the  extra  cost  of  installing  and  maintaining  the 
lighting  system.  It  has  been  said  that  a  man  who  is  obliged 
to  keep  one  eye  on  the  danger  points  of  a  machine  has  only  one 
eye  left  to  operate  it.  This  is  unquestionably  true  and  conse- 
quently a  machine  must  be  made  absolutely  safe.  The  factory 
manager  usually  tries  to  accomplish  this  by  putting  a  guard  rail 

*  See  Electrical  World,  page  319,  Feb.  10,  1912. 


474 


TRANSACTIONS  I.  E.   S. — PART  II 


around  the  danger  points  or  else  enclosing  them  entirely.  This 
seems  about  as  reasonable  as  putting  a  rail  around  a  hole  in  a 
street  without  placing  a  lantern  on  it.  Protected  machines  still 
cause  accidents  and  will  continue  to  do  so  until  the  proper  light 
is  provided  and  the  danger  points  brought  well  into  view.  Acci- 
dents are  becoming  more  expensive  each  year  and  disregarding 
all  humitarian  arguments  an  owner  can  no  longer  neglect  to 
protect  the  operators  from  accidents.  Good  light  is  the  most 
effective  protection  that  can  be  provided  and  only  carelessness  on 
the  part  of  the  employee  will  incur  accident  under  these  condi- 


20  100 
It  90 


ffO 
Z  10 


\SOCTL/0ftT 

ilUk 

^^ 

idmllll 

|!     !  ; 

J 

11 

IIIINIIIIII 

]llm^t~^^S^~-i 

1 

I 

r 

||!! 

jjl 

y^Da&Hnc-ga  1=4=           i    = _!*'-'*»  Lj ;  i  -  j  ■ ' 

H 

||            111  =H             CjLooomess 
j~          |             |            III           Gorve: 
1     §              1911 

|             |             |             III  Retort/ I//omi/>otfon. 

June  July       fiuo.      «5f/>.       Oct.       Clov.      D£c.     </&ri.      fee.      Mq#.     flpfi.       fliqt     Jem 
Chart  showing  accident  and  cloudiness  curves. 

tions.  "It  costs  us  a  lot  of  money,  but  it  has  paid  for  itself  in 
less  than  a  year"  said  one  manufacturer.  What  more  can  an 
owner  want?  Certain  courts  have  held  that  failure  to  illuminate 
danger  points  constitutes  "contributory  negligence."  Germany, 
Austria,  Holland  and  France,  realizing  the  importance  of  good 
lighting  conditions,  have  included  lighting  in  their  codes  for  fac- 
tory inspection  of  health  and  safety.  The  accompanying 
diagrams  shows  that  the  maximum  number  of  accidents  occur 
during  the  time  in  which  artificial  light  is  used.  It  is  interesting 
to  note  that  the  accident  curve  is  almost  a  duplicate  of  the  cloudi- 


ness curve. 


FLEXNER  AND  DICKER:     FACTORY   LIGHTING  475 

In  the  installation  one  must  take  into  account  the  position  of 
the  machines,  the  work  that  is  done,  the  location  of  posts,  the  win- 
dows, and  in  fact  every  condition  which  may  in  some  way 
cause  deep  shadows  and  bad  illumination.  The  scope  of  this 
paper  does  not  allow  us  to  enter  into  any  detailed  account  of 
layouts.  As  stated  in  the  introduction,  we  are  only  attempting 
to  emphasize  the  necessity  and  advantage  of  better  factory 
lighting. 

Good  factory  lighting  is  not  beyond  reach ;  it  is  not  something 
that  one  can  only  wish  for.  It  is  a  material  thing  and  may  be 
had  for  the  asking. 

A  great  many  bad  installations  can  be  made  good  ones  by  two 
inexpensive  methods;  either  re-locating  the  units  and  the  addi- 
tion of  proper  reflectors,  or  in  some  cases  by  replacing  existing 
units  with  some  of  the  modern  efficient  type  now  on  the  market. 
It  is  not  hard  to  show  that  the  new  system  will,  within  a  given 
time,  pay  for  itself,  and  in  a  great  many  cases  save  money  over 
the  operating  and  maintenance  expenses  of  the  old  system. 

Assume  that  the  owner  of  a  factory  depends  solely  upon  the 
profits  of  the  work  his  employees  turn  out.  An  equation  express- 
ing output  must  involve  the  personal  equation  of  the  men  and 
there  must  be  a  certain  personal  efficiency  of  each  man  under 
every  condition  in  which  he  works.  If  a  high  priced  man  is 
placed  under  poor  working  conditions  his  work  will  be  no  better 
than  the  low  priced  man  under  good  conditions.  A  manufacturer 
will  usually  buy  a  labor  saving  device  or  a  machine  with  which 
his  workers  can  turn  out  more  or  better  work,  and  he  will  supply 
his  employees  with  tools  of  the  highest  grade  steel  and  have 
men  to  keep  these  tools  in  the  very  best  condition ;  but  he 
often  absolutely  ignores  the  personal  efficiency  of  the  operator 
and  the  conditions  under  which  he  must  work.  He  does  not 
usually  see  all  the  methods  of  making  the  man  as  perfect  as  his5 
tools.  In  other  words,  more  time  and  thought  is  given  to  the 
tools  than  the  operator.  What  good  is  a  perfect  tool  or  machine 
if  the  operator  can  hardly  see  what  he  is  doing  with  it?  This 
sounds  ridiculous,  of  course,  but  it  is  true  of  many  a  factory 
to-day.  For  instance,  a  manufacturer  purchased  a  certain 
machine  at  a  cost  of  $18,000.00  and  paid  a  high  priced  man  of 


476  TRANSACTIONS  I.  E.   S. PART  II 

long  experience  to  operate  it.  Yet  this  owner  could  not  see  his 
way  clear  to  spend  $19.00  in  order  that  this  high  priced  operator 
would  not  have  to  take  the  product  twenty  feet  away  to  the 
window  to  caliper  it.  This  shows  how  little  the  owner  consid- 
ered the  personal  efficiency  of  his  men. 

The  cost  of  illumination  as  compared  with  an  operator's  sal- 
ary is  very  small  and  insignificant;  in  fact,  so  small  that  the 
manufacturer  can  not  see  it  at  all.  The  following  data,  taken 
as  average  conditions,  shows  this. 

If  a  100-watt  lamp  is  assumed  for  each  man  and  that  it  burns 
3^/3  hours  per  day  for  300  days,  the  following  is  derived : 

Cost  of  lamp  (Commonwealth  Edison  Co.  renewal) $0.00 

Cost  of  reflector 1 .00 

Cost  of  wiring  per  outlet 4.00 

Total  first  cost $5-oo 

Interest  on  investment  6% $0.30 

Depreciation  at  12^  f0 0.63       $0.93 

Power  at  5  c. 5.00 

Cleaning  at  3  c.  per  mo 0.36 

Renewal  of  lamps 0.00 

Total $6.29 

Wages  for  10  hours  a  day,  300  days,  may  be  assumed  to  be 
$1,000.00.  Thus  the  ratio  of  the  cost  of  furnishing  illumination 
to  a  man  under  the  above  conditions  would  be  (overhead  expense 

not  included,),  — : ,  or  0.629  per  cent. 

1,000 

The  following  mathematical  deduction  shows  what  good  light- 
ing would  mean  to  a  factory  upon  the  installation  of  such  a 
system.  Taking  an  area  of  30,000  square  feet  with  an  average 
of  0.75  watt  per  square  foot,  a  connected  load  of  22,500  watts 
would  result.  Figuring  the  installation  with  250  watt  units  an 
estimate  of  the  first  cost  is  surprisingly  low : 

90  250-watt  outlets  at  $3.50 $315.00 

90  fixtures  at  $1.25 112.50 

90  reflectors  at  $1.00 90.00 

90  lamps  (Commonwealth  Edison  Co.  renewal)  • . . 0.00 

Total I5I7-50 

Let  it  be  supposed  that  this  factory  turns  out  a  yearly  business 


FLEXNER  AND  DICKER:     FACTORY   LIGHTING  477 

of  $250,000  and  that  33^3  per  cent.,  or  $83,333.33  of  this  busi- 
ness is  done  under  artificial  light.  Assuming  a  conservatively 
5  per  cent,  increase  in  output  as  the  benefit  due  to  good  lighting, 
the  business  is  then  increased  $4,166.67.  If  there  is  a  profit  of 
20  per  cent,  on  this  output  a  credit  of  $833.33  is  derived,  which 
is  considerably  more  than  the  installation  cost. 

As  further  proof  of  the  low  installation  and  operating  costs  of 
good  lighting  the  following  data  are  submitted  from  a  table  com- 
piled from  actual  figures  on  three  trial  installations  in  a  large 
factory  with  lamp  prices,  etc.,  revised  so  as  to  be  up-to-date. 

100- Watt  Tungsten  Lamp. 

30  reflectors  at  92  c $27.60 

Wiring  at  $3. 22  per  outlet 96.60 

30  lamps  at  0.72 21.60 

Total $i45-8o 

Interest  on  investment  at  6  fo $8. 75 

Depreciation  at  12^  fo 15-55 

Renewals  at  30  X  900/1000  hrs.  X  °-72 *9-44 

Energy  3000  X  9°°  nrs-  X1-1  c- 29.70 

Labor  (cleaning  30  X  0-63  X  20  c. ) 3.78 

Total  annual  cost $77-22 

These  figures  are  derived  on  the  assumption  that  good  factory 
lighting  will  necessitate  a  100-watt  lamp  for  100  square  feet  of 
working  area  required  by  an  ordinary  workman.  With  these 
assumptions  the  following  information  has  been  tabulated : 

Total  working  hours  300  X  10 3000  hours 

Total  lighting  hours  300  X  2>XA IOO°  hours 

Average  cost  of  labor  per  hour 35   cents 

Labor — 

3000  hours  at  35  c. $1050.00 

Light- 
Cost  of  1  oo-watt  tungsten   lamp    (Commonwealth   Edison 

Co.  renewal) $0.00 

Cost  of  metal  reflector  (trade  price) 1.00 

Average,  cost  of  wiring  per  outlet 4.00 

Initial  investment  per  outlet $5-Oo 


47§  TRANSACTIONS  I.  E.  S. — PART  II 

Interest  at  6% 0.30 

Depreciation  at  12^  % 0.63      0.93 

Cleaning  1 2  mo.  at  3  c. 0.36 

Lamp  Renewals  (Maintenance) 0.00      0.36 

Energy  100  K.  W.  H.  at  5  c. 5.00 

Annual  operation  cost 6. 29 

Annual  wages  for  one  man $1050.00 

Cost  of  light  in  per  cent,  of  wages .6 

When  reduced  to  cost  per  hour  based  on  3,000  working  hours 
per  year,  one  finds: 

Labor  per  hour $o-35 

Light  per  hour  - 0.00629 

Cost  of  light  per  day 0.02096 

Cost  of  labor  per  day 3.50 

These  figures  go  to  show  that  the  cost  of  good  lighting  is  a 
very  small  portion  of  the  cost  for  a  man's  time;  in  fact,  if  good 
lighting  would  save  five  minutes  of  a  man's  time  per  day 
a  material  gain  would  be  experienced. 

By  following  this  form,  any  local  conditions  causing  different 
prices  than  those  given  can  be  substituted  so  that  a  comparative 
figure  can  be  obtained  for  any  particular  locality. 

The  cost  of  maintenance  of  tungsten  lamps  and  reflectors  is 
stated  as  follows  in  vol.  1  of  the  191 1  Proceedings  of  the  Na- 
tional Electric  Light  Association. 

Per  cent. 

Renewals  of  lamps 75 

Renewals  of  broken  reflectors 3 

Changing  reflectors  for  washing 16 

Labor  for  washing  reflectors 2 

Additional  indirect  charges 4 

100 

This  data  is  from  experience  with  an  installation  of  between 
7,000  and  8,000  lamps  and  reflectors. 

With  the  available  units,  it  is  impossible  to  pick  out  one  light- 
ing unit  and  say  that  it  can  be  used  for  all  conditions.  There  is 
no  one  cure  for  all  evils.  .  Individual  conditions  enter  into  the 
problem  and  the  resulting  unit  must  be  best  for  the  conditions 
presented.     The  most  important  qualifications  are  the  following: 


FLEXNER  AND  DICKER:     FACTORY  LIGHTING  479 

Efficiency;  color;  quality;  arrangement  of  machines — processes; 
adaptability ;  special  architectural  features ;  and  available  hanging 
height. 

The  best  unit  to  use  will  be  the  one  that  best  fulfills  these 
requirements.  Each  light  source,  whether  gas  arc,  individual 
gas,  electric  incandescent,  arc  or  vapor  lamps,  has  its  definite 
field  in  factory  lighting.  Usually  where  one  should  be  used  the 
others  will  be  less  satisfactory.  It  is  hard  to  convince  the  owner 
that  the  cheapest  is  not  the  best,  for  he  usually  wants  light  only, 
and  often  will  not  pay  for  the  necessary  equipment  to  produce 
illumination.  The  problem  of  which  one  to  use  depends  upon 
the  class  of  work  to  be  done  under  it,  as  each  lamp  has  certain 
characteristics  that  argue  for  and  against  its  use. 

The  last  few  years  have  brought  great  developments  in  the  arc 
lamp.  The  flame  arc  of  long  life,  furnishes  a  light  source  of 
high  candle-power  and  low  maintenance  cost.  When  the  white 
light-giving  carbons  are  used  the  light  emitted  is  of  good  but 
rather  variable  color.  This  lamp  should  never  be  used  in  the 
normal  range  of  vision.  It  is  best  adapted  to  factories  with 
high  ceilings. 

There  has  been  considerable  talk  about  the  harmful  ultra- 
violet rays  emitted  from  arc  lamps.  These  rays  are  no  doubt 
given  off  to  a  considerable  extent,  but  they  are  lost  in  the  inner 
globe.  Therefore  this  characteristic  should  not  be  an  argument 
against  the  arc  lamp.  The  greatest  objection  to  this  light  source 
is  its  unsteadiness,  and  for  fine  accurate  work  a  more  steady  unit 
might  better  be  used. 

The  mercury-vapor  lamps  are  particularly  well  adapted  to 
certain  kinds  of  manufacturing.  The  peculiar  color  together 
with  the  high  visual  acuity  renders  them  very  useful.  A  large 
clothing  manufacturing  concern  has  recently  replaced  enclosed 
arcs  with  vapor  lamps  in  pressing  rooms.  It  is  remarkable 
the  way  scorching  can  be  detected  under  this  lamp  while  if 
a  tungsten  lamp  is  used  the  scorch  is  not  so  noticeable.  The  vapor 
lamp  has  met  with  decided  approval  in  this  kind  of  work.  This 
goes  to  show  that  the  unit  used  should  depend  entirely  upon  the 
work  to  be  done. 


480  TRANSACTIONS  I.  E.   S. — PART  II 

In  installations  where  the  tungsten  lamp  is  the  source  of  light, 
too  much  emphasis  cannot  be  put  on  the  subject  of  cleaning. 
The  manufacturer  would  not  allow  his  operators  to  leave  their 
machines  at  night  without  cleaning  them;  the  floors  are  cleaned 
and  each  morning  the  factory  is  found  in  tip-top  shape.  Why? 
So  that  the  work  may  begin  under  the  best  conditions,  all  work- 
ing toward  an  increase  of  output.  In  other  words,  everything 
but  the  lighting  equipment  is  systematically  taken  care  of.  The 
owner  knows  that  the  time  and  money  spent  in  cleaning  a  machine 
is  well  spent,  and  yet  that  which  has  a  greater  effect  on  the  effi- 
ciency of  the  operator  is  left  to  accumulate  dirt  from  day  to 
day  and  in  many  factories  from  month  to  month. 

In  general,  it  is  best  to  have  the  light  source  as  high  as 
possible  above  the  working  plane.  If  it  is  out  of  reach  of  the 
worker,  he  cannot  handle  it  and  thus  it  will  be  free  of  a  coating  of 
oil  or  other  dirt.  Truly  enough,  certain  machines  require  drop 
cords  in  setting  up  the  work  or  changing  the  dies,  but  few  ma- 
chines actually  need  drop  cords  during  their  operation.  One 
big  railroad  shop  in  Chicago  has  adopted  Cooper-Hewitt  lamps 
for  general  lighting  and  drop  cords  are  checked  as  any  other 
tool.  In  this  way  they  are  taken  care  of  and  are  not  used  except 
when  necessary.  It  has  been  our  experience  that  the  worker 
will  use  a  drop  cord  as  long  as  he  has  one  in  front  of  him. 

The  first  move  for  efficient  lighting  is  general  illumination, 
where  possible,  doing  away  with  the  drop  cord  or  as  above  stated, 
making  the  drop  cord  a  working  tool. 

There  are  many  combinations  of  efficient  lighting  systems;  in 
fact,  it  is  a  subject  of  its  own,  so  that  we  will  not  attempt  a  dis- 
cussion. 

Realizing  the  general  disregard  of  good  lighting  as  a  necessary 
and  important  part  of  the  factory  equipment,  and  not  overlooking 
the  attractive  lighting  load  of  this  class  of  service  the  Common- 
wealth Edison  Co.  of  Chicago  decided  to  make  a  proposition  cov- 
ering lighting  installations  for  factories. 

As  has  been  stated,  the  first  cost  of  the  installation  is  too  often 
the  only  obstacle,  and  therefore  this  company  decided  that  the 
first  way  to  make  such  a  proposition  attractive  was  to  do  away 
with  the  first  cost.     To  insure  most  efficient  operation  the  com- 


JXExner  and  dicker:    factory  ughting  481 

pany  includes  in  this  proposition  the  cleaning  and  renewing  of 
all  fixtures  and  lamps. 

The  Commonwealth  Edison  Company's  proposition  is  as  fol- 
lows: 

The  customer  is  asked  to  sign  a  contract  for  a  period  of 
twenty-four  consecutive  months.  After  the  expiration  of  this 
period,  the  wiring  and  fixtures  become  the  property  of  the  con- 
sumer. 

Charges  for  this  service  are  made  up  on  the  following 
basis :    rental  charge,  maintenance  charge  and  electricity  charge. 

Rental  Charge. — The  rental  charge  is  twenty-five  cents  per 
fixture  per  month,  allowing  the  consumer  to  use  either  100,  150 
or  250-watt  units  in  each  fixture.  At  the  end  of  the  two  year 
period  this  equipment  becomes  the  property  of  the  consumer  and 
this  charge  is  discontinued. 

Maintenance  Charge. — The  consumer  pays  the  company  twen- 
ty-five cents  per  fixture  per  month,  except  during  the  months 
of  June,  July  and  August.  At  the  end  of  the  two  year  period, 
the  consumer  may  elect  to  discontinue  paying  this  charge  and 
take  care  of  this  equipment  himself. 

Electricity  Charge. — For  this  service  the  consumer  pays  our 
regular  schedule  A  rate  which  is  ten  cents  net  per  kilowatt  hour 
for  the  first  thirty  hours  use  of  the  maximum  demand  per  month 
and  five  cents  net  per  kilowatt  hour  for  all  energy  used  in  excess 
of  this  amount. 

The  fixture  supplied  under  this  contract  is  one  that  was  es- 
pecially designed  for  this  class  of  service.  It  consists  of  a 
shallow  reflector  with  a  collar  containing  a  lock  socket ;  the 
conduit  serves  as  a  stem.  The  reflector  is  so  designed  that  the 
filament  does  not  extend  below  the  bottom  of  the  reflector. 
Photometric  curves  show  extensive  characteristics.  The  idea 
throughout  was  to  make  a  reflector  that  was  efficient,  plain,  and 
easily  cleaned. 

The  Commonwealth  Edison  Company  confidently  expects  to 
install  10.000  of  these  fixtures  within  the  next  year,  and  a  re- 
port of  the  first  few  months  gives  reason  for  the  confidence 
expressed. 

In  summing  up  we  believe  that  the  campaign  for  good  fac- 


482  TRANSACTIONS  I.  E.   S. — PART  II 

tory  lighting  has  just  begun  and  that  the  best  argument  in  favor 
of  better  illumination  is  a  statement  showing  the  benefits  derived 
from  an  efficient  lighting  system  and  the  experiences  of  others. 

Even  a  hasty  reconsideration  of  the  arguments  presented  in 
this  paper  demonstrates  the  tremendous  scope  and  possibilities 
along  this  line.  There  is  no  longer  any  excuse  for  poor  light- 
ing; the  necessity,  the  practibility,  and  the  economy  of  good 
illumination  have  been  demonstrated  beyond  question  and  if 
the  strides  in  this  direction  which  have  been  made  in  the  recent 
past  may  be  taken  as  an  index  of  those  which  will  be  made  in 
the  future,  there  is  no  doubt  that  very  soon  the  time  worn  phrase 
"a  badly  lighted  shop"  will  have  disappeared  from  the  vocabulary 
of  those  connected  with  the  lighting  industry. 

We  believe  that  if  a  fair  and  broad-minded  manufacturer  will 
but  figure  out  in  a  common  sense  way  the  merits  and  necessity  of 
good  illumination,  he  will  be  converted  to  its  use  in  a  short  time. 
If  his  own  figures  do  not  satisfy  him,  let  him  consult  those  who 
have  been  far-sighted  enough  to  go  ahead  with  his  better  sense 
dictation  and  be  shown,  if  necessary,  the  truths  of  the  above 
assertions.  He  will  realize  sooner  or  later  the  needs  of  his  men 
— better  atmosphere,  lighter  and  cleaner  shops,  and  proper  illu- 
mination. 

discussion. 

Mr.  H.  A.  Reid  :  On  the  seventh  page  of  the  paper  there 
seems  to  be  a  discrepancy  with  regard  to  the  assumed  cost  of 
wiring  per  outlet,  the  reason  for  which  is  not  apparent.  In  the 
first  table  the  cost  for  100-watt  lamps  is  given  as  $4.00,  and  in 
the  second  table  $3.50  is  used  as  the  cost  for  250-watt  outlets.  It 
would  seem  that  the  cost  of  wiring  should  be  no  higher  for  the 
100-watt  than  for  the  250-watt;  rather  should  the  reverse  be  the 
case.  In  the  actual  installation  cited  on  the  eighth  page  the  cost 
per  100-watt  outlet  was  $3.22. 

The  assumed  depreciation  rates  appear  also  to  vary.  In  the 
first  example  on  the  seventh  page,  12.5  per  cent,  is  taken  to  ap- 
ply on  lamps,  reflectors  and  wiring;  in  the  second  example  the 
rate  on  wiring  is  reduced  to  5  per  cent.  If  the  wiring  is  prop- 
erly installed,  the  latter  value  will  probably  be  about  right. 


FACTORY   LIGHTING  483 

Mr.  R.  B.  Ely  :  I  think  the  central  stations  and  gas  companies 
particularly  the  smaller  companies,  should  encourage  the  intro- 
duction of  illuminating  engineering  departments.  The  up-to- 
date  companies  can  effect  a  saving  and  improve  the  lighting  con- 
ditions in  factories,  stores,  etc.,  and  of  course  where  there  is 
gas  or  other  means  of  illumination  employed,  the  advice  of  the 
engineering  department  when  followed  will  frequently  offset  the 
expenses  of  the  improvements.  In  numerous  instances  it  takes 
a  week  to  lay  out  an  installation.  In  one  case  it  cost  $1,200  to 
equip  one  floor  for  trial,  but  the  effect  and  saving  was  so  good 
that  they  went  ahead  and  wired  the  whole  factory.  That  is  not 
an  exceptional  instance. 

Mr.  G.  H.  Stickney:  I  want  to  emphasize  the  importance 
of  industrial  lighting  to  the  central  station.  I  have  felt  in  the 
past  that  a  great  many  central  stations  have  considered  indus- 
trial lighting  as  an  undesirable  load  and  have  not  awakened  to 
the  possibilities  of  it.  I  think  that  now  that  the  Commonwealth- 
Edison  Company,  one  of  the  most  enterprising  companies  of  the 
country,  has  pointed  out  what  they  are  doing,  we  may  hope  that 
in  another  year  other  central  stations  may  also  contribute  useful 
data. 

Mr.  G.  W.  Roosa  :  In  the  second  paragraph  on  the  tenth  page 
is  this  statement:  "When  the  white  light  giving  carbons  are 
used  the  light  emitted  is  of  good  but  rather  variable  color."  In 
factory  lighting  where  flame  carbon  arc  lamps  are  to  be  used  the 
bays  are  usually  wide  and  high.  Such  installations  often  include 
a  number  of  these  lamps ;  for  instance,  four  or  more,  or  possibly 
a  multiple  of  four.  If  there  is  a  variation  of  candle-power  from 
any  one  lamp  it  can  not  be  altogether  objectionable  because  the 
average  variation  from  several  lamps  will  be  slight,  due  to  the 
fact  that  variations  in  individual  lamps  seldom  occur  simul- 
taneously. Flame  carbon  arc  lamps  are  usually  equipped  with 
diffusing  glassware  of  some  kind  when. placed  within  the  field  of 
vision,  or  at  a  position  approximating  it.  Equipped  with  proper 
diffusing  glassware  this  type  of  lamp  has  an  intrinsic  bril- 
liancy of  about  15  candle-power  per  square  inch.  On  the  other 
hand,  most  incandescent  lamps  are  used  with  clear  bulbs  instead 


484  TRANSACTIONS  I.  E.   S. — PART  II 

of  frosted  ones,  and  I  believe  the  intrinsic  brilliancy  is  about 
1,000  candle-power  per  square  inch. 

Mr.  E.  W.  Lloyd  :  It  occurs  to  me  that  the  benefit  to  be  de- 
rived by  the  manufacturers,  central  stations  and  illuminating  en- 
gineers from  the  development  of  good  lighting  in  factories  would 
put  more  money  in  their  pockets  than  from  any  other  line.  We 
have  heard  a  great  deal  about  the  developments  of  motor  drive 
in  central  station  service  in  factories  all  over  the  United  States. 
The  recent  United  States  census  shows  some  very  remarkable 
figures  relative  to  that  growth — several  hundred  per  cent,  in  10 
years.  I  believe  the  growth  will  be  greater  in  the  next  10. 
Strange  to  relate,  however,  the  development  in  factory  lighting 
has  been  very,  very,  slow ;  at  least  it  is  slow  in  many  of  the  cities 
that  I  am  acquainted  with.  There  are  many  motor  equipment 
in  factories,  but  an  examination  of  the  installations  of  lighting 
in  those  factories  is  a  shock  to  a  man  who  spends  any  time  think- 
ing about  the  subject  of  illumination.  It  was  because  of  that 
condition  in  Chicago  that  we  decided  to  help  increase  this  busi- 
ness by  offering  to  finance  the  installation  of  these  lighting  sys- 
tems. 

Manufacturers  are  becoming  interested  in  the  proper  design 
of  reflectors  and  fixtures,  and  I  believe  there  is  a  great  deal 
being  done  in  this  direction.  All  manufacturers  of  fixtures  are 
not  up  to  date.  Some  of  our  trouble  comes  from  the  different 
schools  of  illumination,  the  different  methods  of  accomplishing 
results;  but  I  don't  think  it  is  necessary  to  dwell  upon  that. 
Every  effort  toward  better  illumination  is  going  to  help  the  whole 
matter  along. 

I  merely  want  to  say  in  closing  that  the  central  stations  can 
do  more  business  at  the  moment,  by  devoting  energy  to  the  illum- 
ination of  factories,  in  my  judgment,  than  in  any  other  way. 

Mr.  H.  C.  Chapin  :  It  is  important  to  emphasize  that  tests 
be  made  of  working  conditions ;  for  instance,  a  yellow  flame 
carbon,  already  of  higher  efficiency  than  a  white  flame,  will  show 
a  still  greater  comparative  efficiency  in  a ,  smoky  atmosphere, 
because  the  yellow  light  penetrates  the  atmosphere  better  than 
the  white. 

Mr.  S.  L.  E.  Rose  :    It  may  be  of  interest  to  know  that  there 


FACTORY  LIGHTING  485 

have  recently  been  developed  reflectors  for  the  flame  arc,  which 
give  intensive  and  extensive  distributions  and  also  an  angle  re- 
flector. A  6.5-ampere  direct-current,  multiple,  flame  lamp  with 
yellow  carbons  and  an  intensive  reflector  gives  about  10,000 
lumens  in  the  zero  to  60  degree  zone.  This  is  equivalent  to  about 
14  lumens  per  watt,  and  applies  where  the  lights  necessarily 
hang  high.  The  extensive  reflector  on  the  same  lamp  gives 
approximately  the  same  number  of  lumens  in  the  zero  to  60 
degree  zone. 

Mr.  M.  H.  FlExnEr:  I  wish  to  say  that  our  information  as 
to  the  increase  in  output,  resulting  from  better  lighting  condi- 
tions, was  obtained  from  a  pamphlet  on  factory  lighting  by 
Mr.  G.  C.  Keech  and  from  the  plant  of  Wilcox  and  Parker  of 
Chicago.  The  latter  firm  reported  about  a  10  per  cent,  increase 
in  output. 

Our  remarks  on  the  flame  arc  lamp  were  based  upon  our 
own  observation.  Flaming  arcs  should  not  be  hung  low,  and  by 
that  I  mean  hanging  heights  of  10  and  12  feet.  In  such  in- 
stallations bright  spots  are  plainly  visible,  proving  that  the  lamps 
are  not  lighting  as  large  an  area  as  they  would  with  higher 
hanging.  The  City  of  Chicago  has  recently  passed  an  ordinance 
which  requires  that  all  arc  lamps  in  the  down  town  district  be 
hung  23  feet,  and  in  the  outlying  districts  21  feet.  There  are 
some  streets  lighted  with  flaming  arcs  and  which  are  hung 
about  10  feet  high.  In  riding  in  the  street  cars  on  those  streets 
one  finds  the  glare  from  these  lamps  very  disagreeable,  and 
riding  through  the  alternate  dark  and  light  spaces,  soon  fatigues 
the  eye.  This,  of  course,  is  a  little  foreign  to  our  subject,  yet 
the  results  would  be  somewhat  similar  in  both  installations  re- 
ferred to. 

The  discrepancy  in  costs  of  wiring  per  outlet,  given  on  the 
seventh  and  eighth  pages,  referred  to  by  Mr.  Reed  is  due  to 
the  fact  that  one  is  an  actual  and  the  other  an  average  cost.  The 
discrepancies  in  depreciation,  etc.,  are  explained  by  the  fact  that 
conduit  was  used  in  one  case  and  moulding  in  another. 

Mr.  H.  H.  Magdsick:     This  paper  brings  to  our  attention 
the  very  low  cost  of  lighting  compared  with  the  total  cost  of 
production    in    industrial   plants.      Recently   we   obtained   com- 
8 


486  TRANSACTIONS  I.  E.   S. — PART  II 

plete  data  for  a  cold-roll  mill  lighted  uniformly  to  an  intensity 
of  two  foot-candles,  which  made  it  possible  to  read  micrometers 
with  ease  in  every  part  without  any  auxiliary  drop  lamps.  The 
total  cost  of  operating  the  lighting  units,  including  the  fixed 
charges  on  the  equipment,  renewal  and  maintenance  of  lamps 
and  the  energy,  is  only  one  one-hundred-fiftieth  of  the  cost  of 
labor  during  the  hours  of  darkness.  In  other  words,  supply- 
ing all  the  light  that  is  desired  in  this  department  costs  no 
more  than  furnishing  one  helper  for  each  150  men.  The 
difference  in  cost  between  poor  illumination  and  adequate  light- 
ing is,  of  course,  still  less.  Any  system  that  would  enable 
this  mill  to  be  operated  at  all  would,  we  will  say,  require  one- 
half  the  expense;  the  difference  between  poor  lighting  facili- 
ties and  the  very  best  illumination  is,  then,  secured  at  a  cost 
no  greater  than  that  of  supplying  one  helper  for  300  men. 
And  the  cost  of  labor  is  only  one  item  in  the  cost  of  pro- 
duction, often  no  more  than  20  per  cent,  of  the  total.  If  to 
it  are  added  the  fixed  charges  on  the  plant  and  machinery,  the 
cost  of  power  and  expense  of  supervision,  administration,  etc., 
one  obtains  an  idea  of  how  comparatively  insignificant  is  the 
cost  of  proper  illumination. 

The  iron  and  steel  industry  offers  an  interesting  example  of 
recent  developments  in  plant  illumination.  President  Parkhurst 
of  the  Association  of  Iron  and  Steel  Electrical  Engineers  has 
given  permission  to  quote  from  a  paper  that  will  be  presented 
before  the  convention  of  that  association  now  in  session.  The 
data  contained  therein  were  compiled  from  reports  submitted 
by  nearly  one-half  of  the  companies  represented  in  the  A.  I.  & 
S.  E.  E.  It  will  be  recalled  that  the  members  have  for  several 
years  interested  themselves  actively  in  illumination  matters  and 
were  among  the  first  to  co-operate  with  this  society. 

From  the  reports,  it  appears  that  during  the  past  two  years 
the  wattage  installed  in  lighting  equipment  throughout  the  plants 
has  decreased  slightly,  about  5  per  cent,  but  the  average  inten- 
sity of  illumination  has  been  raised  to  about  135  per  cent,  of 
the  former  value.  This  increase  was  secured  by  replacing  obso- 
lete equipment  with  more  modern  units.  Thus,  while  two  years 
ago  about  58  per  cent,   of  the  connected  lighting  load  was  in 


factory  lighting  487 

carbon  arc  lamps,  for  the  most  part  of  the  220-volt  type,  and 
less  than  8  per  cent,  was  in  high-efficiency  incandescent  lamps, 
the  carbon  arc  lamps  have  now  dropped  to  about  34  per  cent, 
and  the  high-efficiency  incandescent  units  have  been  increased 
to  an  equal  wattage.  The  carbon  incandescent  lamps  have  not 
been  decreasing  as  rapidly  as  the  arcs  due,  it  appears,  to  the 
fact  that  up  to  the  present  time  the  intensities  of  general 
illumination  provided  have  proved  insufficient  to  warrant  dis- 
continuing the  use  of  drop  lamps  for  local  lighting.  Where 
complete  modern  systems  have  been  installed,  however,  the 
reports  show  that  auxiliary  units  have  in  practically  all  cases 
disappeared.  These  changes  represent  the  important  develop- 
ments in  the  lighting  equipment  of  the  steel  mills.  The  flame 
arcs,  mercury-vapor  lamps,  etc.,  remain  a  small  percentage  of 
the  total. 

According  to  the  data  submitted,  the  76  plants  represented 
in  the  A.  I.  &  S.  E.  E.  still  have  about  24,000  kilowatts 
connected  in  carbon  arc  and  carbon  incandescent  lamps.  If 
these  were  replaced  with  modern  equipment,  the  illumination 
instead  of  being  135  per  cent,  of  the  average  value  of  two  years 
ago,  would  become  235  per  cent.  It  is  costing  the  plants  about 
$1,500,000  annually  to  operate  this  obsolete  equipment.  At  a 
conservative  estimate,  fully  two-thirds  of  this  amount,  or 
$1,000,000,  is  wasted  annually  in  view  of  the  higher  efficiency 
of  the  modern  equipment  with  which  it  might  be  replaced. 
The  fact  that  when  changes  are  made  this  wasted  amount  will 
probably  be  expended  in  the  operation  of  modern  lighting  sys- 
tems in  order  to  secure  higher  intensities  of  illumination,  serves 
merely  to  multiply  the  loss,  for  the  gain  in  the  output  of  the 
mills  and  increased  safety  of  employees  is  in  practically  every 
instance  far  greater  than  the  cost  of  illumination. 


488  TRANSACTIONS  I.  £.  S. — PART  II 

HOSPITAL  LIGHTING.* 


BY  WIWJAM   S.    KILMSR. 


Synopsis:  Short-comings  in  hospital  lighting  are  too  well  known  to 
need  mention;  ordinary  illumination  devices  for  hospitals  fail  to  give  the 
proper  quality  and  distribution  of  light.  This  paper  treats  solely  of  the 
lighting  of  the  two  most  important  quarters  of  the  hospital,  viz.,  operating- 
room  and  wards,  and  describes  practical  fixtures  for  the  solution  of  the 
various  problems.  Although  written  from  an  engineering  standpoint  the 
paper  is  based  on  intimate  knowledge  of  hospital  conditions  derived  from 
the  author's  two  years'  concentrated  work  in  this  field. 

The  modern  hospitals  of  to-day  are  undoubtedly  models  of  ele- 
gance and  hygienic  forethought  and  this  high  degree  of  perfec- 
tion has  been  reached  by  experience  in  overcoming  specific  diffi- 
culties and  conditions  as  they  are  encountered. 

For  some  unexplainable  reason  the  question  of  proper  artificial 
illumination  has  received  little  or  no  consideration,  thus  greatly 
lowering  the  efficiency  of  the  splendid  service  characteristics  of 
these  institutions. 

Light  like  any  other  form  of  energy  may  become  an  agent  of 
destruction  or  a  minister  of  health,  precisely  in  accordance  with 
the  wisdom  shown  in  its  application,  and  it  is  the  duty  of  the 
professional  advisor,  whether  he  be  architect  or  engineer,  to 
understand  these  particular  conditions  before  he  can  pretend 
to  satisfactorily  specify  a  lighting  system. 

The  operating  rooms  and  wards  which  are  the  most  important 
parts  of  a  hospital  naturally  present  the  most  difficulties,  and  it  is 
the  object  of  this  paper  to  treat  particularly  the  lighting  of  these 
two  areas. 

OPERATING  ROOMS. 

The  operating  table  should  have,  on  account  of  the  large  amount 
of  surgical  work  performed  at  night,  a  system  capable  of  con- 
centrating on  the  field  of  operation  an  illumination  intensity  of 
not  less  than  25  foot-candles,  preferably  higher,  and  approximate 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Societ3',  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


KILMER:     HOSPITAL  LIGHTING 


489 


the  tonal  value  of  daylight.  The  equipment  should  be  designed 
along  plain  smooth  lines,  thus  avoiding  the  septic  risks  of  dust 
collections  and  rendering  it  easily  and  thoroughly  cleaned.  When 
it  is  necessary  to  suspend  a  fixture  in  close  proximity  to  the 
table  and  when  over  250  watts  are  consumed  by  the  equipment, 


Fig.  1. — Ventilated  operating  fixture,  using  six  ioo-watt  tungsten  lamps. 

it  is  most  important  that  the  heat  generated  by  the  lamp  be 
directed  away  from  the  patient  and  the  head  of  the  surgeon. 

Fig.  1  illustrates  a  unit  of  the  above  description,  which  is  de- 
signed for  six  1 00- watt  tungsten  lamps  and  when  suspended 
6  ft.  6  in.  (1.98  m.)  from  the  floor  to  lower  edge  of  the  frame, 


49Q 


TRANSACTIONS  I.  E.   S. — PART  II 


an  average  illumination  of  40  foot-candles  is  distributed  over  the 
table.  By  the  addition  of  a  ball  and  socket  joint  this  equipment 
may  be  adjusted  to  suit  any  form  of  operation. 

To  reduce  the  heat  generated  by  the  lamps  to  a  minimum,  the 
fixture  has  an  ingenious  arrangement  of  double  glass  slides,  by 


Fig.  2. — Ventilated  operating  fixture,  using  eight  35-watt  double  base  tubular  lamps, 
with  emergency  gas  attachment. 

which  a  forced  draft  is  created  by  the  heating,  and  consequently 
raising  the  air  between  the  glasses,  the  heated  air  passing  out  of 
the  fixture  through  the  vents  at  the  top.  This  syphon  arrange- 
ment is  often  assisted  by  the  use  of  suction  pumps. 

Under  actual  working  conditions,  temperature  tests  show  the 


KILMER:     HOSPITAL  LIGHTING 


491 


following  results,  with  mercury  9  in.  (22.86  cm.)  below  the  lower 
plate  glass : 

Degrees 
Fahrenheit 

Temperature  of  room 7& 

After  one  hour  burning  with  slide 83 

After  one  hour  burning  with  no  slide 94 

Fig.  4  shows  an  equipment  of  this  character  installed  in  the 
Post  Graduate  Hospital,  New  York  City. 


Fig.  3. — Detail  of  dustless  indirect  ward  fixture. 


Fig.  2  is  a  unit  built  on  the  same  general  principles,  but  designed 
for  a  double  base  tubular  lamp  with  a  straight  filament.  This 
fixture  with  eight  35-watt  lamps  gives  an  illumination  equal  to 
the  previously  described  unit,  but  over  a  smaller  area.  The 
elimination  of  the  excessive  heat  is  as  follows : 

Degrees 
Fahrenheit 

Temperature  of  room •  •   78 

Fixture  after  one  hour  burning  with  slide 81 

Fixture  after  one  hour  burning  witii  no  slide 87 


492  TRANSACTIONS  I.  E.   S. — PART  II 

In  both  of  these  fixtures  the  mirrors  forming  the  reflecting 
surface  are  so  arranged  that  the  field  in  front  of  the  surgeon's 
hand  is  always  free  from  shadows,  no  matter  in  what  position 
it  may  be. 

The  new  concentrated  filament  lamp,  with  a  distribution  of 
light  given  in  the  following  table,  has  made  possible  another 
form  of  lighting  the  operating  table,  vis.,  adjustable  reflectors 
capable  of  powerful  concentration.  (Electrical  Testing  Labora- 
tories report  on  ioo-watt  concentrated  tungsten-filament  lamp 
referred  to)  : 


Angles 

Candle- 

degrees 

power 

155 

66.O 

145 

72.0 

135 

75-o 

125 

75-5 

115 

76.0 

105 

76.5 

95 

74.0 

90  Horizontal 

74.0 

85 

76.0 

75 

78.0 

65 

80.0 

55 

81.0 

45 

82.0 

35 

78.0 

25 

72.0 

15 

64.0 

5 

56.0 

0  Nadir 

55-o 

Fig.  6  is  an  all  metal  reflector  so  designed  that  the  light  is  con- 
fined approximately  to  an  area  50  either  side  of  the  vertical  and 
should  be  suspended  stationarily  directly  over  the  table,  about 
8  ft.  (2.44  m.)  from  the  floor.  When  it  is  so  installed  with  a 
1 00- watt  concentrated  filament  lamp  a  satisfactory  distribution  is 
given  for  all  operations  requiring  a  powerful  downward  distri- 
bution. To  protect  the  eye  of  attendants  and  surgeons,  when 
raised  from  the  field  of  operation,  a  series  of  metal  bands  may 
be  inserted  as  shown  in  the  drawing. 

The  following  photometric  values  apply  to  this  unit: 


^ 


Fig.  3a.  — Exteiior  of  dustless  indirect  ward  fixture. 


Fig.  4.—  Fixture  shown  in  Fig.  1  installed  over  operating  table  in 
Post  Graduate  Hospital,  New  York  City. 


Fig.  5. — Adjustable  parabolic  reflector  for  powerful  concentration. 


Fig.  6. — Concentrating  reflector  for  use  directly  over  operating  table. 


KILMER:     HOSPITAL  LIGHTING  493 

Apparent  Candle-power 

Angles ' 1 

Degrees  Without  bands  With  bauds 

o  Axis 2,160  2,160 

5  2,250  j, 800 

10  430  380 

Photometric  distance  30  ft.  (9.144  m.)  ;  measurements  made  in 
a  single  plane  through  the  reflector  axis. 

Abdominal  and  pelvic  operations  require  a  penetrating  beam 
of  light.  Fig.  5  shows  a  parabolic  all  metal  reflector  18^5  in- 
(47.59  cm.)  in  diameter,  adjustable  to  any  angle.  This  equip- 
ment should  be  attached  to  a  rigid  support  about  12  ft.  (3.66  m.) 
from  the  floor  and  12  ft.  to  15  ft.  (3.66  to  4.57  m.)  to  the  rear 
of  the  table  with  the  adjusting  rod  of  sufficient  length  to  be  easily 
reached;  when  so  installed  a  powerful  beam  of  light  is  directed 
over  the  shoulder  of  the  surgeon  directly  into  the  field  of 
operation. 

The  photometric  values  of  this  unit  are  as  follows : 

Angles                                                                                     Apparent  Candle-power 
Degrees  , — : • , 

0.0 9,600  7.940 

2-5 8,330  — 

5.0 2,000  — 

Photometric  distance  30  ft.  (9.144  m.)  ;  100-watt  concentrated 
filament  tungsten  lamp,  measurements  made  in  a  single  plane 
through  the  reflector  axis. 

It  is  quite  apparent  that  with  this  distribution  there  need  be 
no  fear  of  insufficient  and  poorly  distributed  light.  These  equip- 
ments may  also  be  mounted  on  adjustable  standards,  thus  having 
an  equipment  for  all  emergencies  which  are  so  often  encountered. 

Some  hospitals  object  to  any  suspended  lighting  equipment  in 
the  operating  room;  this  usually  applies  to  rooms  having  a  sky- 
light, and  provided  sufficient  head  room  is  available  a  satisfac- 
tory system  may  be  designed  by  placing  a  series  of  reflectors 
around  the  edge  of  the  skylight.  These  reflectors  may  be  so 
designed  that  the  maximum  reflection  falls  at  a  point  directly 
over  the  table.  This  scheme  is  likely  to.be  more  satisfactory  than 
an  attempt  to  illuminate  the  entire  skylight  so  as  to  provide  a 
sufficient  amount  of  light  for  an  operation  in  any  part  of  the 
room.  The  latter  scheme  would  necessitate  a  very  large  con- 
sumption of  energy. 


494  TRANSACTIONS  I.  £.   S. — PART  II 

WARD  LIGHTING. 

In  the  wards  the  evils  of  glare  have  to  be  most  carefully 
avoided,  as  influence  of  glare  upon  the  retina  of  the  debilitated 
or  depressed  is  usually  followed  by  serious  results. 

The  average  ward  of  twenty  beds  should  be  provided  with  two 
kinds  of  illumination,  general  and  localized.  The  general  light- 
ing should  not  average  more  than  one  foot-candle  and  should  be 
obtained  from  either  a  totally  or  partially  indirect  source,  so 
arranged  that  violent  contrasts  are  entirely  eliminated.  It  is 
advisable  that  the  ceiling  be  treated  with  a  flat  white  or  French 
zinc  surface  and  the  walls  in  a  slightly  darker  tone,  such  as 
buff  or  French  gray. 

The  fixtures  must  be  of  special  design  and  the  same  hygienic 
conditions  apply  as  in  the  operating  room,  vis.,  elimination  of 
dirt  and  heat.  Also  that  portion  of  the  light  which  is  diverted 
to  the  ceiling  must  be  evenly  distributed  without  spots  and 
streaks.  The  exterior  finish  should  be  a  restful  color  such  as 
green  enamel  or  matt  nickel.  This  will  not  give  a  violent  con- 
trast, and  will  overcome  the  sameness  of  the  white  enamel  finish 
which  many  fixture  manufatcurers  seem  to  think  is  necessary 
for  hospitals. 

Fig.  3  shows  a  form  of  indirect  fixtures  which  has  met  with 
great  success  in  ward  ligthing.  It  can  be  made  for  any  num- 
ber of  lamps  up  to  six.  The  interior  is  lined  with  a  series  of 
mirrors  set  at  any  angle  to  insure  an  even  distribution  of  light 
over  the  ceiling.  The  angle  at  which  these  are  set  depends  on 
the  distance  between  outlets  and  the  size  of  the  ward.  Over  the 
top  of  the  bowl  is  set  a  thin  blown  glass  plate  which  may  be 
easily  raised  for  lamp  renewal  by  means  of  the  sliding  metal 
holder,  and  which  renders  the  interior  of  the  fixture  dust  and 
dirt  proof.  Ventilation  is  accomplished  by  means  of  vent  holes 
at  the  bottom  where  the  air  enters  and  passes  through  and  es- 
capes between  the  housing  of  the  fixture  and  the  glass  plate.  A 
view  of  these  fixtures  installed  in  one  of  the  wards  of  the  Post 
Graduate  Hospital,  New  York,  is  shown  in  Fig.  7. 

If  it  is  so  desired  one  lamp  in  a  fixture  may  be  arranged 
to  operate  on  a  separate  switch,  but  this  is  not  advisable  on 
account  of  its  effect  on  the  patients.     Night  lights,   wherever 


/ 


VI H 


Kig.  7. — Night  view  of  typical  ward  in  Post  Graduate  Hospital.  New  York, 
using  dustless  fixture  shown  in  Figs.  3  and  3a. 


ft 


^_ 


0 


~y^y~ 


Figs.  8  and  8a.— Adjustable  ward  bracket,  lower  figure  showing 
correct  method  of  installing. 


Fig.  9. — Conical  microscopic  reflector. 


KILMER:     HOSPITAL  LIGHTING  495 

necessary  should  be  in  the  form  of  brackets  and  as  inconspicu- 
ous and  as  far  out  of  the  range  of  vision  as  possible. 

When  a  direct  indirect  system  is  to  be  used  the  character  of 
the  glassware  used  for  the  bowl  or  hemisphere  must  receive  a 
thorough  study ;  it  should  be  of  sufficient  density  to  maintain  the 
presently  given  intensities  per  square  inch  of  exposed  sources  and 
have  diffusion  qualities  which  render  it  impossible  to  have  any 
bright  spots  or  streaks  from  the  lamps  exposed  to  the  eyes  of 
the  patients.  It  is  also  advisable  to  eliminate  the  necessary  gas 
emergency  equipment  from  these  fixtures,  but  it  may  be  included 
in  the  localized  equipment. 

For  the  localized  lighting  of  the  ward  the  two  forms  most 
advisable  are  proper  brackets  or  "bedside"  lights ;  everything 
depends  on  the  word  "proper."  No  source  of  light  or  illuminated 
surface  exceeding  0.05  candle-power  per  square  inch  (7.2  candle- 
power  per  square  foot)  should  be  exposed  to  a  patient's  eye; 
therefore,  the  usual  direct  lighting  equipment  is  quite  out  of  the 
question.  An  excellent  form  of  bracket  fixture  is  shown  in  Fig. 
8.  It  is  adjustable  in  every  direction  and  it  is  equipped  with  a 
reversible  glass  reflector  with  opaque  sides.  The  reflecting  por- 
tion is  of  a  concentrating  type,  designed  especially  for  bedside 
examination,  thus  eliminating  any  necessity  of  the  use  of  a  special 
portable  lamp  by  attending  physicians ;  and  by  reversing  directly 
upward  this  concentrated  light  is  directed  to  the  ceiling,  thus 
giving  an  excellent  form  of  indirect  illumination. 

The  opaque  sides  render  it  impossible  for  a  ray  of  light  to 
reach  the  eye  of  a  patient  in  any  part  of  the  ward  and  enable 
any  bed  to  be  immediately  illuminated  without  any  disturbance  to 
other  patients.  An  entire  ward  equipped  with  one  of  these 
brackets  to  each  bed  would  not  require  any  other  form  of  illum- 
ination and  the  emergency  gas  equipment  may  be  easily  incor- 
porated in  this  fixture.  Many  hospital  equipment  specialists  and 
authorities  predict  this  to  be  the  future  form  of  ward  lighting. 

"Bedside"  lights  are  the  least  desirable  for  a  localized  equip- 
ment, but  where  absolutely  necessary,  they  should  be  carefully 
designed.  The  reflector  portion  should  preferably  be  hemis- 
pherical in  form,  opaque  and  mounted  on  a  plain  standard 
heavily  weighted  at  the  base.    The  reflector  should  be  adjustable 


496  TRANSACTIONS  I.  E-   S. — PART  II 

by  means  of  a  knuckle  joint.  The  reflecting  surface  should  give 
a  diffused  reflection  as  from  opal  glass  or  flat  white  enamel. 
Of  course  an  additional  light  is  required  for  the  physician  for 
bedside  examinations,  thus  making  this  equipment  more  costly 
in  both  initial  expense  and  maintenance,  as  both  equipments  have 
to  obtain  the  energy  from  base  board  receptacles  which  naturally 
necessitate  a  considerable  portion  of  exposed  cord,  making  the 
liability  of  breakage  considerable  and  costly. 

In  the  examination  of  bacteria  under  a  powerful  microscope, 
it  is  always  difficult  to  arrive  at  definite  results  when  observers 
at  various  points  are  working  under  light  sources  of  varying 
spectral  value.  Varying  light  sources  naturally  produce  a  dif- 
ferent tonal  value  to  a  germ  so  that  it  is  often  difficult,  if  not 
impossible,  to  be  sure  of  identification. 

Fig.  9  shows  an  instrument  whereby  there  may  be  obtained 
at  any  time  a  uniform  light  source  giving  approximately  the 
effect  of  a  slightly  veiled  north  light,  which  is  so  much  desired 
in  precision  work,  and  far  superior  to  the  average  illumination 
obtainable  from  city  windows.  The  light  from  a  60-watt  all 
frosted  tungsten  lamp  operating  at  exact  line  voltage  is  directed 
by  means  of  a  double  convex  reflector  with  the  lower  surface 
of  silvered  glass  and  the  upper  surface  of  opal  glass. 

A  carefully  calibrated  liquid  lens  is  inserted  at  the  aperture. 
After  the  light  has  passed  through  this  lens  the  resultant  illum- 
ination on  the  microscopical  field  is  of  a  daylight  spectral  value. 
For  dark  field  work  this  lens  is  changed  to  one  giving  an  illum- 
ination of  a  powerful  violet  character. 

Other  lighting  apparatus  of  the  hospital  has  been  intention- 
ally not  discussed  in  this  paper  because  in  the  author's  opinion 
it  does  not  call  for  the  study  given  the  conditions  herein  men- 
tioned. 

The  lighting  of  halls  and  corridors  is  in  a  class  with  the 
lighting  of  hotels  and  office  building  corridors.  The  nurses  and 
doctors  quarters  should  receive  the  same  careful  treatment 
required  for  the  lighting  of  the  home.  The  dispensary  should 
be  given  the  same  treatment  as  the  modern  drug  store;  particu- 
lar attention  should  be  given  to  the  lighting  of  the  prescription 
counter. 


HOSPITAL  LIGHTING  497 

DISCUSSION. 

Mr.  H.  Calvert:  This  paper  is  a  good  article  for  a  person 
to  read  who  is  about  to  design  the  lighting  of  a  hospital.  It  is 
very  practical.  The  lamp  fixtures  shown  for  the  operating  room 
certainly  would  give  good  results ;  but  one  reason  why  the  author- 
ities of  some  hospitals  object  to  having  the  lamps  placed  directly 
over  the  operating  table  is  the  liability  of  dust  or  dirt  falling 
from  the  fixture  and  infecting  the  patient.  To  obviate  this,  some 
hospitals  have  adopted  the  plan  of  installing  four  units  approxi- 
mately above  the  corners  of  the  operating  table. 

Referring  to  ward  lighting,  indirect  fixtures  are  quite  suitable 
provided  the  amount  of  illumination  which  falls  on  the  ceiling 
is  not  excessive,  as  the  patients,  a  large  part  of  the  time,  are 
lying  on  their  backs  and  to  gaze  upwards  at  a  very  bright  ceiling 
is  of  course  uncomfortable.  The  adjustable  ward  bracket  which 
is  shown  in  the  paper  is  a  good  one  for  the  purpose.  In  one  of 
the  Philadelphia  hospitals  this  bracket  arrangement  has  had  a 
unique  feature  added  by  equipping  the  bracket  with  a  pull  socket 
so  that  the  patient  can  light  the  lamp,  and  at  the  same  instant  a 
little  pilot  lamp  which  is  located  near  the  desk  of  the  head  nurse 
is  also  lighted  indicating,  if  she  has  not  already  observed  it, 
that  the  patient  desires  attention. 

Mr.  W.  F.  Little  :  Referring  to  the  lighting  of  operating 
tables,  I  note  that  Mr.  Wheeler  states  that  an  intensity  of  10- 
foot-candles  has  been  found  sufficient.  Comparing  this  with  il- 
luminations given  by  the  Zeiss*  system,  we  find  intensities  in 
the  neighborhood  of  300-foot-candles,  and  I  further  understand 
that  this  system  has  been  used  with  very  good  success  in  various 
hospitals.  I  should  be  interested  in  knowing  whether  Mr. 
Wheeler  believes  10-foot-candles  sufficient  for  operating  table 
lighting  in  general. 

Admittedly,  the  light  of  hospital  wards  should  be  done  under 
the  principle  of  the  greatest  good  for  the  greatest  number. 
Therefore,  would  it  not  be  feasible  to -turn  an  indirect  system 
of  illumination  upside  down  directing  the  light  first  upon  the 
white  floors  and  from  there  diffused  throughout  the  room? 

*  Illumination  measurements  of  Zeiss  system  given  in  I.  E.  S.  Transactions,  June, 
1912. 


498  TRANSACTIONS  I.  E.   S. — PART  II 

Mr.  H.  B.  WhEeX£r:  Operating  Rooms.  The  lighting  of 
hospitals  is  a  very  important  subject  and  one  we  know  very  little 
about.  This  is  especially  true  in  operating  room  illumination. 
It  has  been  my  experience  that  extreme  concentration  of  light 
on  the  operating  table,  not  only  gives  a  multiplicity  of  shadows, 
but  an  intense  heat,  both  the  opposite  to  what  is  desired,  namely, 
an  evenly  diffused  light.  Indirect  Illumination  from  the  ceiling 
and  walls  gives  an  abundance  of  diffusion  eliminating  completely 
objectionable  sharp  shadows. 

Several  large  operating  rooms  of  the  Toronto  General  Hos- 
pital and  St.  Mary's  Hospital  at  Rochester,  Minn.,  have  been 
lighted  for  some  time  with  indirect  illumination  which  is  giving 
very  satisfactory  results  for  operating  purposes.  The  indirect 
fixtures  are  very  plain  arm  chandeliers  with  adapters  for  sup- 
porting one  piece  silvered  glass  reflectors.  The  average  inten- 
sity is  approximately  10-foot-candles. 

Another  system  of  diffuse  illumination  has  been  in  use  for 
the  past  six  years  in  the  Southern  Pacific  Hospital  at  Los  An- 
geles, Cal.  It  consists  of  a  battery  of  one  piece  silvered  glass 
reflectors  suspended  over  a  chipped  glass  skylight.  In  installing 
a  system  of  this  character,  care  should  be  taken  to  select  glass 
of  good  diffusing  qualities. 

Corridors.  Corridor  lighting  is  just  as  important  as  the  light- 
ing of  any  other  room  in  a  hospital,  because  patients  are  con- 
tinually being  taken  from  the  wards  and  private  rooms  to  the 
operating  table  at  critical  stages.  For  this  reason  the  lighting 
should  be  concealed  as  in  other  parts  of  the  institution. 

Wards.  It  is  very  desirable  in  wards  to  have  a  flexible  sys- 
tem of  indirect  illumination. 

A  low  intensity  {%.  foot-candle)  for  a  night  light,  a  medium 
intensity  (y2  to  2  foot-candles)  for  reading,  etc.,  and  a  high  in- 
tensity for  close  examination  of  patients  are  required.  Local- 
ized direct  lighting  portables  attached  to  baseboard  outlets  be- 
tween beds  are  used  largely  for  the  high  intensity. 

Generally  in  large  wards  varying  degrees  of  illumination 
are  controlled  by  electrolier  wall  switches  and  the  smaller  wards 
by  switches  on  the  fixtures. 

When  luminous  bowls  are  desired,  the  small  lamp  for  illumi- 
nating the  glass  bowl  may  be  used  for  a  night  light. 


philbrick:   store;  lighting  499 

STORE  LIGHTING.* 


BY    J.    E.    PHILBRICK. 


Synopsis:  The  gaslighting  installations  of  eight  small  stores  are  out- 
lined in  this  paper.  Plans  showing  the  locations  of  test  stations,  and 
illumination  readings  of  each  store  are  also  included.  The  author  em- 
phasizes two  points:  (i)  gaslighting  solicitors  should  have  available  for 
prospective  customers  accurate  data  on  local  lighting  installations;  (2)  the 
need  of  proper  maintenance  to  insure  maximum  efficiency  and  satisfaction. 

The  author  hopes  that  the  publication  of  the  tests  and  data 
of  this  paper  and  a  discussion  of  them  may  go  far  towards  elim- 
inating the  old  "hit  or  miss"  methods  used  by  commercial  depart- 
ments in  securing  business  and  also  in  making  installations. 

The  following  table  gives  average  resultsf  of  tests  of  the  light- 
ing installations  in  several  stores,  plans  of  which  are  shown  in 
the  accompanying  illustrations. 

Table  1. 

lumens  per 
cu.  ft.  of  gas  per 
hour  light  ceiling 

Nominal  eon-  With  With 

sumption  of  light  dark 

fighting  unit  unit  cu.  ft.  walls  walls 

Reflex  lamps  with  frosted  tip  cylin- 
ders, prismatic  or  light  (imported) 
opal  concentrating  reflectors t>1A  125  114 

Reflex  lamps  with  frosted  tip  cylin- 
ders, prismatic  or  light  (imported) 
opal  distributing  reflectors 3^  no  100 

Reflex  lamps  with  frosted  tip  cylin- 
ders, French  roughed  ball  globe-..       t>xA  95  70 

Reflex  cluster  lamp,  four-mantle,  with 
alabaster  globe 13  S5  64 

Inverted  five-mantle  arc  with  alabaster 

globe 16.6  87  65 

Upright  four-mantle  arc  with  opal  re- 
flector and  alabaster  globe 20  75  55 

Upright  four-mantle  arc  with  alabaster 
globe  only 20  66  48 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating   Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 

t  Figures  published  by  a  manufacturer  of  lighting  appliances  for  use  in  the  design  of 
gas  lighting  installations. 


500  TRANSACTIONS  I.  E.   S. — PART  II 

The  installations  tested  were  selected  at  random  and  under 
ordinary  working  conditions,  no  special  preparation  of  lamps  or 
mantles  being  made  in  any  case.  If  renewals  had  been  made  they 
were  made  at  the  ordinary  time  for  cleaning,  and  as  the  test 
shows  there  is  but  one  case  in  which  the  units  had  received  atten- 
tion within  one  day  of  the  time  of  making  tests. 

For  some  time  the  opinion  has  prevailed  that  figures  relative 
to  the  performance  of  gas  lighting  units  required  radical  dis- 
counting in  order  to  express  the  actual  results  secured  upon  a 
consumer's  premises,  under  the  conditions  of  care  and  attention 
usually  encountered  in  actual  service. 

Obviously,  it  is  most  important  for  the  man  responsible  for 
the  success  of  any  business  to  obtain  accurate  information  con- 
cerning the  excellence  of  the  service  rendered  by  his  product 
with  particular  reference  to  those  features  directly  under  the 
observation  of  the  consumer,  and  upon  which  the  latter  bases 
his  judgment  of  the  service  and  product.  Many  gas  companies 
have  been  deterred  from  gathering  this  data  by  the  effort  re- 
quired to  obtain  it.  I  believe,  however,  that  the  value  of  this 
information  is  amply  demonstrated  in  the  tests  reported  in  this 
paper  which  were  made  in  York,  Pa.  Tests  of  this  character 
not  only  enable  the  manager  to  determine  the  competitive  posi- 
tion of  his  product  and  to  keep  a  check  on  the  capacity  of  his 
manufacturing  distribution  and  maintenance  departments  as  re- 
gards the  performance  of  their  various  functions  in  supplying 
lighting  service,  but  form  a  basis  for  calculation  by  salesmen  in 
designing  installations  and  advising  customers. 

When  salesmen  use  data  furnished  by  the  manufacturer  of 
lighting  appliances,  they  are  naturally  inclined  to  make  rather 
liberal  discounts  to  allow  for  discrepancies  between  service  and 
laboratory  conditions,  and  for  the  pardonable  optimism  naturally 
to  be  expected  on  the  part  of  the  manufacturer. 

The  accompanying  table  published  in  the  "Gas  Solicitor's 
Handbook"  was  believed  at  the  time  of  publication  to  be  repre- 
sentative of  the  results  that  might  be  expected  in  actual  service 
under  good  conditions  with  clean  lamps  and  new  mantles  of 
good  quality,  and  laboratory  tests  have  indicated  that  the  de- 
preciation which  results  from  burning  over  the  reasonable  period 
of  time  elapsing  between  maintenance  calls  should  be  negligible. 


philbrick:    store  lighting  501 

It  was  desired  to  substantiate  this  in  practise,  and  1  was  very 
much  gratified  to  find  that  this  was  done. 

Lacking  facilities  for  determining  the  gas  consumption  in  each 
case  2>TA  cubic  feet  per  hour  was  taken  as  presumably  a  close 
approximation  to  the  actual  consumption,  the  sizes  of  the  mantles 
indicating  that  on  gas  of  the  quality  furnished,  this  assumption 
was  reasonably  accurate. 

Tests  made  upon  new  mantles  of  known  efficiency  confirmed 
this  opinion. 

In  determining  upon  a  procedure  for  these  tests,  horizontal 
illumination  upon  the  working  plane  was  selected  as  the  basis. 
Not  because  this  is  the  only  plane  requiring  illumination  but 
because  with  practically  all  the  glassware  usually  sold  for  store 
illumination,  a  sufficient  degree  of  horizontal  illumination  is 
always  accompanied  by  at  least  a  sufficiency  of  illumination 
upon  other  planes.  The  main  purpose  was  to  obtain  information 
which  might  assist  us  in  furnishing  illumination  to  our  customers 
under  the  most  favorable  conditions. 

There  are  several  matters  that  are  worthy  of  some  special 
attention  in  connection  with  these  different  installations.  It  will 
be  noticed  that  in  many  cases  the  illumination  was  measured  at 
comparatively  few  points,  and  while  the  numerical  average  of 
the  results  would  not  give  the  actual  average  obtained  throughout 
the  entire  area,  the  numerical  average  is  in  all  probability  some- 
what below  the  true  average  which  would  have  been  obtained 
had  a  greater  number  of  readings  been  taken.  For  instance,  in 
both  Figs.  1  and  2,  it  will  be  seen  that  most  of  the  measurements 
were  taken  in  the  more  poorly  illuminated  portions  of  the  room, 
very  few  being  taken  at  points  immediately  beneath  the  lamps 
where  the  illumination  would  be  the  highest.  This  statement  also 
applies  to  Figs.  3,  4  and  5.  With  regard  to  Fig.  7  the  actual 
average  illumination  is  probably  far  above  the  figures  which  we 
have  given,  possibly  as  much  as  20  to  30  per  cent.,  on  account 
of  the  fact  that  the  lamps  were  quite  low  and  most  of  the  read- 
ings were  taken  at  situations  so  far  from  the  lamps  that  they 
were  outside  of  the  range  of  effective  distribution.  In  this  par- 
ticular instance,  of  course,  the  really  important  consideration  is 
the  amount  of  light  on  the  face  of  the  customer,  and  the  lamps 
9 


502 


TRANSACTIONS  I.  E.   S. — PART  II 


are  properly  arranged  to  distribute  the  light  for  this  purpose  in 
the  best  manner,  and  the  general  illumination  of  the  room  is 
of  much  less  importance. 


/ 


6 

If      »: 

Is ! 

t 

.-If. 

.J 

* 

g 

|s 

Hh\/ 


o  _ 


In  the  store  shown  in  Figs.  7  and  8,  it  will  be  noticed  that  a 
greater  number  of  readings  were  taken  and  these  probably  rep- 
resent truer  averages  than  any  of  the  others. 

An  interesting  point  is  that,  according  to  the  Gas  Solicitor's 
Handbook,   (page  30),  is  claimed  that  a  cubic  foot  of  gas  per 


philbrick:    store  lighting 


503 


hour  will  produce  about  no  effective  lumens  in  a  room  with 
light  side  walls  (that  is,  a  sufficient  amount  of  light  to  illuminate 


'Uiili-ftif 
Fig.  3. — Plan  of  shoe  store. 


a 

1 

1 

r« 

«E 

a: 
u 
t 

z 

0 

0 

COUNTER 

05 

a 

04 

o3 

-8 

Ot 

ov 

n 

■t 

Ol 

a 

SCALE  OF  FT 


Fig.  4. — Plan  of  jewelry  store. 


SCHLC  OF  TUT 

Fig.  5. — Plan  of  pool  room. 


0* 

* 

OT 

»• 

■  , . 

9 

m 

- 

.. 

0* 

„ 

1 10  square  feet  to  an  intensity  of  1  foot-candle)  and  with  dark 
walls  100  effective  lumens.  From  the  numerical  averages,  as  ob- 
tained in  the  test,  it  will  be  noted  in  the  last  column  that  there 


504 


TRANSACTIONS  I.  E.   S. — PART  II 


was  only  one  gas  installation  which  dropped  a  sufficient  degree 
below  the  nominal  efficiency  stated  in  the  handbook  to  make  the 
matter  worth  considering,  and  this  is  the  barber  shop,  in  which 
as  I  stated  above,  the  lamps  were  so  low  that  most  of  the  readings 
came  outside  of  the  effectively  lighted  area,  so  that  as  far  as  in- 
dicating the  efficiency  of  the  gas  service  and  maintenance,  this 
should  be  eliminated. 

The  installation  shown  in  Fig.  I  had  not  been  maintained  for 
two  months,  and  was  only  i  per  cent,  below  the  nominal  efficiency. 
Installation  No.  2  was  3  per  cent,  below. 


□ 


CISAR 
COUNTER 


ml 


on   0 


«»i  q     °* 


m 


A 


2T    X      X      a 


I I I I 1- 


iCAit  in  rtrr. 


Fig.  6. — Plan  of  barber  shop. 


The  value  of  tests  of  this  character,  both  to  the  gas  company 
and  to  the  industry  at  large,  is  to  my  mind  quite  apparent  and 
extends  not  only  to  the  solicitation  of  new  business,  but  to  the 
proper  design  of  lighting  systems  and  to  the  maintenance  of 
existing  installations. 

As  regards  the  soliciting  of  new  business,  actual  tests  of  in- 
stallations of  which  the  prospect  has  personal  knowledge  rein- 
forces the  arguments  of  the  salesman  most  forcibly. 

In  many  cases,  contracts  for  lighting  hinge  mainly  upon  econo- 


PHILBRICK  :     STORE   LIGHTING 


505 


mic  considerations.  Every  salesman  claims  the  highest  economy 
for  the  particular  illuminant  he  happens  to  be  selling.  His  claims 
are  of  necessity  based  upon  laboratory  tests,  or  upon  service  tests 
in  other  localities  in  which  the  conditions  may  or  may  not  ap- 
proach those  in  his  own  situation.  He  can  submit  no  evidence 
of  the  validity  of  his  claims  that  has  much  weight  with  the  cus- 
tomer. In  such  cases  tests  upon  installations  in  the  same  locality 
are  very  convincing,  particularly  if  some  of  them  happen  to  have 
been  made  upon  the  premises  of  the  customer. 


/ 


wrwv    c 


~/ 


/■■ 


w 


I     I     I   T~T 


I'    I"  f  "I      1 


fl 


Fig.  7. — Plan  of  shoe  store. 


£ 


I         '       ! 


TTTCL!;  M-'= 


Fig.  8.— Plan  of  carpet  and  rug  store. 

In  the  case  of  the  barber  shop  quoted  above,  the  proprietor 
himself  took  simultaneous  readings  with  the  operator  and  saw 
for  himself  the  true  comparison  between  competitive  illuminants 
so  far  as  the  amount  of  light  he  was  getting  was  concerned,  and 
even  though  the  amount  of  light  fell  43  per  cent,  below  the  theo- 
retical, the  illumination  was  nearly  twice  that  which  was  given 
when  the  shop  was  lighted  by  the  competitive  illuminant. 

A  test  like  this  gives  the  consumer  an  exact  idea  of  the  com- 
parative value  of  competitive  illuminants  and  makes  a  valuable 
reference  for  the  solicitor  of  the  gas  company,  inasmuch  as  the 
conditions  are  approximately  the  same  under  any  city  and  the 
solicitor  has  but  to  refer  to  the  test  and  the  consumer  personally 
for  a  confirmation  of  the  data  with  which  he  is  trying  to  secure 
the  business  of  a  prospective  customer. 


506 


TRANSACTIONS  I.  E.   S. — PART  II 


Photometric  Data. 


Fig.  i. 


Fig.  2. 


Station 

Foot-candles 

Station 

I 

1.6 

I 

2 

2.6 

2 

3 

3-8 

3 

4 

4.0 

4 

5 

4-3 

5 

6 

4.1 

6 

7 

3-7 

7 

8 

2-7 

8 

9 

2-3 

9 

IO 

5-2 

10 

II 

7.8 

11 

12 

6.8 

12 

13 

7.0 

*3 

14 

6.7 

14 

15 

6.0 

15 

16 

3-65 

16 

17 

1.6 

17 

18 

3-2 

18 

19 

4-5 

19 

20 

4-5 

20 

21 

4.6 

22 

4-5 

23 

4-9 

24 

2.4 

Foot-candles 
6.0 

5-3 

4-5 

3-3 

4-5 

5-2 

4.2 

1.85 

5-8 

5-3 

3-7 

1.7S 

5-i 
5-o 
4.0 
2.2 
5-6 

3-3 

1.74 

2.2 


Fig-  3- 


Fig.  4- 


Station 

Foot-candles 

I 

3-1 

2 

3-9 

3 

5-7 

4 

7-3 

5 

4.8 

6 

6.6 

7 

2.25 

8 

5-8 

9 

5-i 

10 

2.0 

11 

6.4 

12 

4.0 

13 

6-3    - 

14 

4-4 

15 

3-2 

16 

3-° 

Station 

Foot-candles 

I 

2.2 

2 

5-5 

3 

7-8 

4 

7.0 

5 

4.0 

6 

i-9 

7 

4.25 

8 

8.0 

9 

7.2 

10 

4-7 

PHILBRICK  :     STORE    LIGHTING 


507 


Fig.  5- 


Fig.  b. 


Station 
I 

Foot-candles 
I.30 

Station 
I 

Foot-can  lies 
3-6 

2 

I. OO 

2 

6.2 

3 

4 

I.l6 
.80 

3 

4 

3-5 
2-5 

5 
6 

.90 
13.00 

5 
6 

5-6 
3-o 

7 

6.50 

7 

4.0 

8 

12.50 

8 

7-3 

9 

5-50 

9 

i-5 

10 

9.20 

10 

4-7 

11 

12 

I.46 
.86 

11 
12 

7.0 
2.8 

13 

1.77 

13 

3-3 

14 

*-i3 

14 

4.4 

15 

1.70 

15 

1-45 

Fi 
Station 
I 

g-  7- 

Foot-candles 

2.20 

Station 

I 

Fig.  8. 

Foot-candles 

i-34 

2 

3.00 

2 

2.09 

3 

4-5o 

3 

2-74 

4 
5 
6 

5.00 
3.60 
2.60 

4 
5 
6 

2.40 
LIS 

1-34 

7 

4-3o 

7 

1.96 

8 

5-15 

8 

309 

9 

4- 25 

9 

3-05 

10 

3.00 

10 

1-52 

11 

2.30 

11 

1.70 

12 
13 

14 

3.60 
6.08 

5.26 

12 
13 
14 

3-°3 
3-53 
3-47 

15 
16 

3-54 
3.20 

15 

1.47 

17 
18 

4.45 
♦      5-4o 

19 

4.20 

20 

3.10 

21 

2.70 

22 

3.00 

23 

3.65 

24 

3-75 

25 
26 

2.90 
2.20 

• 

27 

2S 

3-55 
4.10 

29 

3-55 

SO 

2.40 

5o8 


TRANSACTIONS  I.  t.   S. — PART  II 


Gas 


Fig. 
No. 

I 

2 

3 
4 
5 
6 

7 


Fig. 
No. 

I 

2 
3 

4 

5 
6 

7 
8 


Area  Ceiling 

Business  sq.  ft.            height 

Tailor  shop 979         i4/-o// 

Lunch  room 2,704         \2'-&f/ 

Shoe  store 2,066         i2/-o// 

Jewelry  store 450         io/-6// 

Poolroom 1,487           y'S" 

Barber  shop 611           8'-&> 

Shoe  store 1,916         i5/-o// 

Carpet  store 2,31s         i8/-o// 

Lamps  per 

Business  outlet 

Tailor  shop 2 

Lunch  room 14-2L.,  3-1L 

Shoe  store 2-1L.,  1-2L., 

6-3L.,  2-4L. 

Jewelry  store 2 

Pool  room 4 

Barber  shop 2 

Shoe  store 2 

Carpet  store 3 


Height  of 
lights 

9'- 

Walls 
Medium 

No.  of 
outlets 

6 

io'-6" 

Medium 

17 

$'-6" 

Light 

11 

&-&' 

Dark 

4 

7'-8" 

Dark 

5 

y/_2// 
io'-o" 

Light 
Light 

6 
10 

il'-o" 

Dark 

6 

Gas 

Type 

6  Reflex 

Reflector 
Extensive  prismatic 

Fig. 
No. 

I 

2 

3 
4 
5 
6 

7 


Total  con. 
Business  per  hour 

Tailor  shop 40.0  Cu.  ft. 

Lunchroom 102.3    " 

Shoe  store 99.0   " 

Jewelry  store 26.4   " 

Pool  room 66.0    " 

Barbershop 39.6   " 

Shoe  store 66.0   " 

Carpet  store 59.0    " 


Illumination 


Average       Minimum    Maximum 


Eff.  I,m.  per 
cu.  ft.  or  watt 

Fig.  1 ' . 

No.                    Business  Actual          Theor. 

i         Tailor  shop 104.0            105 

2  Lunch  room 108.0            105 

3  Shoe  store 96.2            no 

4  Jewelry  store 90.0            100 

5  Pool  room 88.3            100 

6  Barbershop 62.6            no 

7  Shoe  store 109.5             no 

8  Carpet  store 87.6             100 

*  Increase 


4.26 

T 

6 

7-8 

4.2 

I 

74 

6.0 

4.61 

2.0 

7-3 

5-25 

I 

■9 

8.0 

3-92 

O 

8 

13.0 

4.06 

I 

45 

7-3 

3-76 

2 

.20 

6.08 

2.25 

I 

15 

3-53 

Gas 

Per  cent. 

Actual 

below 

Time  since 

theor. 

maintained 

O.I 

2  Months 

*3-Q 

I  Day 

12-5 

Not  maintained 

10.0 

1  Month 

12.0 

1  Month 

43-o 

1  Week 

0.5 

1  Week 

12.2 

1  Month 

PHILBRICK  :     STORE   LIGHTING  509 

I  believe  that  it  would  be  of  much  value  to  the  commercial  de- 
partments of  all  the  gas  companies  to  have  tests  like  the  above 
made  and  printed  and  copies  given  to  their  solicitors,  thus  enab- 
ling them  to  meet  the  question  of  the  consumer  "How  do  you 
know"  with  data  which  applies  to  the  question  at  hand  and  not 
an  irrelevant  mass  of  figures  which  mean  nothing  under  the 
local  conditions. 

The  great  feature  of  help  to  solicitors  is  not  that  the  consumer 
will  understand  terms  of  illumination  intensity  but  that  he  can 
see  for  himself  the  real  value  of  his  lighting  and  the  effect  on  his 
pocket-book  in  dollars  and  cents  which  to  him  is  a  thing  of  vital 
interest. 

Heretofore  the  designs  of  installations  have  been  made  at  ran- 
dom, the  lighting  effects  have  often  been  not  satisfactory  upon 
the  first  trial,  a  second  and  sometimes  the  third  attempt  being 
necessary  to  give  the  consumer  the  proper  light  in  the  proper 
place.  This  often  entails  considerable  expense  to  the  consumer 
and  company  and  creates  dissatisfaction. 

It  is  a  fact  that  the  lighting  engineer  of  the  manufacturer  has 
been  at  the  disposal  of  the  gas  companies  for  sometime  back, 
offering  to  lay  out  installations  and  perform  all  the  illuminating 
engineering  work,  but  how  many  of  the  companies  have  taken  ad- 
vantage of  this  offer.  If  solicitors  would  make  a  series  of  tests, 
and  compute  a  table  of  lighting  efficiency  in  various  store-rooms 
they  could  easily  prevent  mistakes  in  future  installations. 

Manufacturers'  hand-books  are  all  right,  but  they  are  too 
often  discounted  or  ignored,  but  data  gotten  right  in  the  field 
under  the  local  conditions  cannot  be  discounted  or  ignored  but 
are  convincing  facts. 

There  is  one  test  which  I  did  not  have  time  to  prepare  but 
would  have  been  advisable  to  undertake  and  that  is  a  test  on 
temperature  at  various  points  in  our  stores  under  the  gas  and 
electric  lighting,  taking  into  account  the  outside  temperature  dif- 
ference on  the  different  days  of  test.  I  believe  the  heat  objection 
against  gas  lighting  can  be  greatly  diminished,  if  not  eliminated. 

Lastly,  it  is  realized  that  the  efforts  of  the  illuminating  engineer 
are  entirely  lost  if  the  lamps  are  not  kept  clean  and  the  mantles 
renewed  at  proper  intervals.     It  is  absolutely  necessary  to  have 


5IO  TRANSACTIONS  I.  E.   S. — PART  II 

men  who  are  careful  and  intelligent,  who  have  some  knowledge 
of  the  lamps  they  are  cleaning;  and  who  know  when  the  adjust- 
ment of  the  lamp  after  cleaning  is  as  near  right  as  possible  to  get 
maximum  efficiency.  Too  often  do  companies  try  to  economize 
on  this  most  important  part  of  their  business  by  employing  boys 
at  very  small  fixed  wages,  and  expecting  them  to  have  the  incli- 
nation to  become  illuminating  experts.  To  my  mind  the  only 
way  to  pay  lamp  maintainers  is  on  the  sliding  scale  plan  with 
the  deduction  for  complaints  on  lamps  which  they  have  maintained 
and  caused  the  consumer  trouble.  The  lamp,  whether  gas  or 
electricity,  must  have  clean  glassware  and  bulbs  for  much  business 
is  lost  by  our  failure  to  attend  to  these  features.  Of  course,  it 
costs  more  to  maintain  lamps  properly  than  it  does  to  clean  them 
in  a  half-hearted  manner,  trying  to  keep  the  cost  down  at  the 
expense  of  the  illumination.  This,  however,  can  be  met  by  a 
proper  and  not  excessive  charge  based  on  the  sliding  scale  of 
consumption  per  mantle.  (It  is  obviously  not  fair  that  the  large 
consumer  of  gas  or  other  illuminants  should  pay  at  the  same  rate 
for  his  maintenance  as  the  consumer  who  uses  his  lights  merely 
as  a  makeshift.)  Below  is  the  maintenance  schedule  which  is  in 
use  by  a  gas  company  in  York,  Pa.,  which  was  adopted  after  care- 
ful consideration  by  the  officers  of  the  company.  This  schedule  is 
giving  entire  satisfaction  to  the  consumer  and  company  alike. 

YORK  GAS  COMPANY 


York,  Pa.,    1913. 

The  undersigned  at  number    Street, 

York,  Pa.,  hereby  makes  application  to  the  York  Gas  Company  to  use  gas 
for  illumination,  at  the  regular  rate  of  the  Company. 

It  is  further  understood  that  this  application,  being  approved,  the  York 
Gas  Company  agrees  to  loan  and  install  the  necessary  Gas  Lighting  Fixtures, 
Piping,  etc. ,  free  of  cost. 

In  all  cases  where  combination  Gas  and  Electric  Fixtures  are  specified 
and  installed,  the  undersigned  agrees  to  use  Gas  regularly  on  such  fixtures 
from  September  1st  to  May  1st,  that  said  fixtures  remain  in  his  premises. 

These  fixtures  will  be  given  regular  monthly  inspection,  new  mantles  sup- 
plied as  needed,  be  cleaned  and  adjusted  and  be  given  such  other  additional 
attention  as  may  be  necessary  to  keep  fixtures  in  good  working  order. 

In  consideration  of  the  above  the  undersigned  agrees  to  pay  maintenance 
according  to  the  following  scale: 


PHIEBRICK  :     STORE   LIGHTING 


5" 


Consumption  per  mantle  per  month, 


(<         « < 


50  feet  or  less 

100    ' 

<     11     << 

125    ' 

1          II            II 

150    ' 

1          <1           II 

175    ' 

•          II           << 

200    ' 

I          II           II 

225    ' 

'          "            " 

250    ' 

I          11            II 

275    ' 

I          II           II 

3°°    ' 

1          II           II 

400    ' 

1          11           11 

500    ' 

1          II           II 

600    ' 

1          II            II 

700    ' 

I          II            11 

800    ' 

I          II           II 

1,000    ' 

1          II           11 

1,250    ' 

1          II           II 

1,500    ' 

III           II 

2,000    ' 

1          II            II 

•15 

cents 

per  month 

•IO 

" 

" 

" 

■    9*2 

" 

11 

" 

•  9 

" 

" 

" 

•  sy2 

" 

" 

it 

.  8 

1 1 

11 

" 

•  7H 

" 

" 

1 1 

■   7 

" 

" 

" 

■  6/2 

" 

" 

" 

.  6 

" 

" 

it 

•   5% 

11 

" 

1 1 

•   5 

" 

" 

1 1 

•  AlA 

" 

" 

" 

•  4 

<i 

" 

•' 

•   3*A 

" 

' ' 

" 

•  3 

it 

" 

" 

•     2/2 

" 

" 

" 

•    2 

" 

1 1 

" 

•  •     I 

11 

11 

All  maintenance  to  be  paid  monthly. 
Orders  taken  by Signed  ■ 

Finally,  gas  men  should  appreciate  the  possibilities  of  gas  light- 
ing, determine  exactly  the  intensity  of  light  in  store-rooms, 
factories,  etc.,  approach  consumers  with  more  intelligible,  con- 
vincing data,  and  keep  the  lamps  clean  and  properly  adjusted. 

By  doing  this,  it  has  been  said,  "Two  units  can  be  made  to 
grow  where  only  one  grew  before." 

The  writer  wishes  to  acknowledge  his  indebtedness  to  Mr.  R. 
F.  Pierce,  for  his  help  in  the  preparation  of  this  paper. 

DISCUSSION. 

Mr.  R.  F.  Pierce  :  Inasmuch  as  these  statistics  were  taken 
for  a  purely  commercial  purpose  and  intended  to  set  forth  only 
the  existing  lighting  service  conditions  in  actual  installations,  no 
deductions  into  which  enter  such  considerations  of  gas  pressure 
or  electric  voltage  should  be  drawn.  While  these  data  were 
taken,  they  were  not  presented  in  the  paper  which  is  simply  a 
statement  of  lighting  conditions  in  a  certain  locality  under  the 
conditions  found,  and  having  no  reference  whatever  to  any  other 
set  of  conditions. 

I  think  Mr.  Philbrick  rather  over-estimates  the  amount  of  as- 
sistance that  he  received  from  me.  I  furnished  the  illumination 
measurements,  and  in  that  connection  should  like  to  explain  the 


512  TRANSACTIONS  I.  E.   S. — PART  II 

method  used.  The  measurements  were  comparatively  few,  the 
idea  being  to  obtain  some  sort  of  an  average  figure  which  should 
at  least  be  low.  It  was  impossible  to  devote  the  time  necessary 
to  take  a  larger  number  of  measurements  which  would  have 
given  more  accurate  figures,  but  the  results  were  comparative 
rather  than  absolute.  Mr.  Philbrick  wished  particularly  to  ob- 
tain an  idea  concerning  his  position  in  the  competitive  field  and 
to  ascertain  how  closely  the  published  figures  for  effective  lumens 
per  cubic  foot  were  approached  in  his  situation. 

In  this  connection,  I  would  emphasize  that  the  table  given  on 
the  first  page  of  the  paper  does  not  refer  to  measurements  ob- 
tained at  York,  but  to  figures  for  effective  lumens  per  cubic  foot 
published  by  a  manufacturer  of  gas  lighting  appliances  for  use 
in  the  design  of  installations.  These  figures  represent  averages 
of  a  number  of  practical  installations  and  have  been  given  out 
as  a  practical  basis  for  the  design  of  gas  lighting  systems. 

In  the  tests  at  York  a  number  of  installations  were  selected 
in  which  combination  fixtures  were  used  with  the  same  type  of 
reflector  on  both  gas  and  electric  outlets  and  measurements  were 
made  on  both  systems  at  the  same  points  to  show  a  comparison 
between  the  two.  The  gas  company  had  no  prior  knowledge 
either  of  the  date  of  the  test  or  the  installations  that  would  be 
tested.  The  latter  were  picked  out  at  random  without  reference 
to  the  condition  of  the  lamps  or  of  the  installation.  The  tests  on 
the  gas  lighting  installation  showed  a  maximum  discrepancy  be- 
tween the  data  published  by  the  manufacturer  and  the  results 
obtained  under  these  tests  of  about  12.5  per  cent.,  with  the  single 
exception  of  one  installation  in  which  the  number  of  test  sta- 
tions was  so  small  that,  on  account  of  the  very  low  height  of  the 
lamps,  the  majority  of  the  test  stations  fell  outside  of  the  effective 
radius  of  the  lamps.  In  this  installation  the  discrepancy  between 
published  figures  and  the  average  obtained  was  43  per  cent. 
That  this  discrepancy  was  due  to  the  location  of  the  stations  is 
indicated  by  the  fact  that  a  test  of  the  electrical  installation  at 
the  same  points  showed  a  discrepancy  of  60  per  cent,  from  the 
commonly  used  data  for  effective  lumens  per  watt,  so  that  in 
both  cases  the  discrepancy  appears  to  be  due  to  the  condition 
mentioned.  I  do  not  mean  to  convey  the  impression  that  the 
relative  service  conditions  found  in  this  situation  exists  in  the 


STORE  LIGHTING  513 

majority  of  plants.  As  I  remember  the  figures,  the  average  dis- 
crepancy for  effective  lumens  per  watt  obtained  in  the  electrical 
installations  was  in  the  neighborhood  of.  40  to  45  per  cent.  It 
is  quite  obvious  that  this  was  due  to  a  large  extent  to  failure  to 
replace  the  lamps  at  proper  intervals  and  to  select  lamps  of 
proper  voltage.  A  comparison  between  the  two  services  em- 
phasizes the  advisability  of  central  stations  and  gas  companies 
obtaining  actual  illumination  measurements  in  order  to  deter- 
mine the  lighting  service  really  given  to  their  customers  under 
the  conditions  of  actual  use.  The  electric  service  in  this  particu- 
lar situation  was  plainly  below  the  average,  while  the  gas  lighting 
service  was  much  better.  There  are,  of  course,  many-  cases  in 
which  this  comparison  would  be  reversed.  The  results  of  this 
test  emphasize  the  fact  that  no  general  comparison  between  the 
service  rendered  by  the  two  illuminants  is  possible.  The  most 
important  factor  is  the  quality  of  service  rendered  in  each  par- 
ticular situation. 

Mr.  T.  J.  LitlE,  Jr.  :  It  is  very  agreeable  to  note  in  Mr. 
Philbrick's  paper  that  the  data  that  he  has  obtained  at  York  com- 
pares very  favorably  with  the  published  data  on  the  same  types 
of  lamps  which  he  tested.  It  is  a  fact  that  published  data  on 
many  of  these  lamps  is  in  many  cases  conservative  from  the 
fact  that  the  tests  were  made  under  2.5  inches  water  pressure, 
while  in  many  cities  the  service  pressure  is  much  higher  than 
this,  and  they  consequently  get  increased  candle-power  and  effi- 
ciency from  their  burners. 

For  instance,  the  consumer's  service  pressure  in  Chicago  is 
6  inches  while  in  San  Francisco  it  runs  considerably  higher  than 
this.  It  may  be  interesting  for  you  to  know  that  considerably 
higher  efficiencies  than  those  which  have  already  been  published 
are  expected  in  the  near  future.  They  have  already' been  reached 
experimentally.  For  instance,  it  is  quite  possible  with  certain 
sizes  of  inverted  lamps  to  exceed  30  candles  per  cubic  foot, 
mean  lower  hemispherical  rating.  In -fact,  I  have  seen  as  high 
as  53  candles  per  cubic  foot,  mean  lower  hemispherical  rating, 
the  lamp  burning  on  low  pressure  {2l/2  inches  water  pressure). 

It  is  a  remarkable  fact  when  looking  back  over  the  develop- 
ment of  gas  and  electric  lighting  systems  that  in  the  year  1880  the 


514  TRANSACTIONS  I.  E.   S. — PART  II 

most  radical  developments  took  place,  namely,  the  Edison  in- 
candescent electric  lamp  and  the  Welsbach  incandescent  gas 
lamp.  Ever  since  that  date,  whenever  there  was  a  radical  de- 
velopment in  the  one  system,  there  was  something  brought  out 
in  the  other  to  meet  it. 

Mr.  Ward  Harrison  :  Mr.  Philbrick's  paper  is  very  inter- 
esting to  me.  It  shows  what  an  effective  weapon  poor  service 
on  the  part  of  one  lighting  company  may  afford  its  competitor. 
It  has  been  found,  however,  that  comparative,  tests  of  this 
character  too  frequently  lead  to  bad  feeling  between  the  lighting 
companies  interested.  This  feeling  can  be  changed  to  one  of 
good  natured  rivalry  simply  by  making  it  a  practise  in  each  case 
to  notify  everyone  interested  in  the  test  and  arranging  for  them 
to  attend.  It  has  been  found,  also,  that  when  the  prospective  pur- 
chaser and  the  representatives  of  all  the  competing  interests  are 
present  at  the  time  of  the  tests,  much  of  the  dissatisfaction  in 
regard  to  test  methods  and  results  which  so  often  follows,  is 
obviated. 

Mr.  E.  B.  Rowe  :  I  cannot  emphasize  too  strongly  Mr.  Har- 
rison's contention  that  operating  companies  should  know  what 
results  the  customers  are  getting  from  service  on  their  lines.  The 
making  of  installation  tests  similar  to  those  covered  in  Mr. 
Philbrick's  paper  is  certainly  to  be  commended,  but  the  condi- 
tions under  which  test  is  made  should  be  clearly  recognized.  In 
making  use  of  such  test  results  complete  data  should  be  given 
or  the  results  should  be  reduced  to  a  common  equivalent  basis. 

For  instance,  on  the  third  page  no  reference  is  made  to  a 
measurement  of  the  gas  pressure,  which  of  course  is  one  of  the 
variables  in  the  test  and  has  a  considerable  influence  on  the  re- 
sults obtained.  In  the  same  way  in  tests  on  an  installation  of 
electric  lamps  the  exact  voltage  on  which  lamps  were  operated 
during  the  test  should  be  stated  and  the  effect  of  voltage  varia- 
tions made  clear  to  the  layman  or  the  results  could  be  reduced 
to  the  normal  voltage  or  gas  pressure.  Personally  I  have  some 
hesitancy  in  making  use  of  published  data  unless  all  the  facts 
are  at  hand  and  I  believe  care  should  be  taken  in  presenting  test 
results  to  cover  all  questions  which  are  likely  to  arise  in  the 
minds  of  those  who  may  wish  to  use  the  data. 


LAW  AND  POWELL:     STORE  LIGHTING  515 

DISTINCTIVE  STORE  LIGHTING.* 


BY   CLARENCE   L.   LAW   AND  A.   L.   POWELL. 


Synopsis:  Certain  stores,  namely  the  high-class  shops,  demand  a 
striking  individuality  of  design.  This  paper  describes  in  detail  several 
typical  lighting  installations  which  have  come  to  the  attention  of  the 
authors;  it  discusses  a  particular  store  of  each  of  several  classes:  shoe, 
millinery,  toy,  candy  stores,  etc.  Data  are  given  as  to  the  dimensions, 
wall,  ceiling  and  floor  coverings,  arrangement  of  fittings,  type  of  glass- 
ware, number  and  sizes  of  lamps,  and  a  general  description  of  the  appear- 
ance of  each  store.  This  information  may  not  be  directly  applicable  to 
the  design  of  a  new  installation,  but  since  the  lighting  systems  outlined 
are  giving  satisfactory  service,  the  quantitative  element  may  be  of  use 
and  the  novelty  of  some  of  the  equipments  may  suggest  to  the  designing 
engineer  ideas  which  will  be  applicable  to  his  particular  problem. 

If  one  pauses  to  analyze  retail  places  of  business  as  a  whole, 
it  becomes  evident  that  a  convenient  and  complete  classification 
may  be  made  as  follows :  ordinary  small  stores,  large  dry  goods 
and  department  stores,  and  high  grade  shops. 

The  authors  treated  the  first  of  these  classes  in  a  paper  pre- 
sented at  the  last  convention1  of  the  Society;  they  realized  that 
high  efficiency  of  light  utilization,  low  initial  cost  of  installation, 
low  maintenance  and  simplicity  were  the  determining  factors.  It 
seemed  desirable  to  suggest  a  standard  practise,  and  an  attempt 
was  made  to  do  this  on  the  basis  of  averages  obtained  from  an  in- 
vestigation of  a  large  number  of  stores  of  this  class.  Some 
criticism  was  elicited  to  the  effect  that  this  scheme  would  pro- 
duce a  monotonous  condition,  but  the  fact  still  remains  that 
artistic  appearance  cannot  be  had  cheaply,  and,  due  to  the  small 
profit  earned  by  a  store  of  this  sort,  the  amount  spent  for  light- 
ing is,  of  necessity,  small. 

In  the  large  store,  efficiency  and  artistic  appearance  become 
more  nearly  balanced.  Artistic  lighting  implies  good  diffusion ; 
and  with  the  present  commercial  illumi'nants,  this  cannot  be  had 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 

1  Transactions  of  the  I.  E.  S.,  Vol.  VII,  p. . 


5l6  TRANSACTIONS  I.  E.   S. — PART  II 

without  some  absorption  of  light.  The  merchant  can  afford  to 
spend  a  relatively  larger  sum  for  lighting  than  could  be  spent 
by  the  small  storekeeper,  and  some  sacrifice  of  light  is  made  to 
obtain  better  diffusion.  The  store  should  have  a  harmonious 
system  of  lighting  for  the  main  parts  of  the  entire  building,  yet 
there  are  some  parts  which  are,  in  reality,  shops,  and  should  be 
so  treated.  The  general  requirements  for  department  store 
lighting  have  been  discussed  several  times  in  the  Transactions* 
of  the  Society,  and  there  is  no  need  for  their  repetition  in  this 
paper. 

A  high  grade  shop  should  be  considered  quite  differently  from 
the  monotonous  store  which  is  so  common  in  large  cities.  Un- 
like a  small  store,  it  should  be  considered  individually  and  with 
respect  to  its  particular  line  of  business.  This  shop  is,  as  a  rule, 
small,  handsomely  and  lavishly  furnished,  splendidly  finished 
to  the  minutest  detail,  and  located  in  the  most  fashionable  sec- 
tion; it  handles  only  the  best  grade  of  goods  (frequently  im- 
ported) and  sells  to  a  discriminating  class  of  customers.  The 
proprietor  or  manager  is  willing  to  spend  large  sums  for  the 
right  equipment  and  maintenance.  The  profits  for  each  indi- 
vidual piece  of  merchandise  sold  are  undoubtedly  greater  than  in 
other  stores,  and  therefore,  more  money  can  be  spent  for  indi- 
viduality of  equipment.  Artistic  appearance  is  the  predominant 
factor,  and,  therefore,  a  distinctive  system  of  lighting  is  neces- 
sary, efficiency  of  the  installation  being  a  secondary  consideration. 

It  should  be  the  aim  and  desire  of  shopkeepers  of  this  class 
to  interest  and  attract  prospective  customers,  making  them  per- 
manent habitues  of  their  stores.  Some  definite  architectural 
scheme  should  be  carried  out  or  symbolism  expressed.  Many 
stores  show  the  influence  of  the  personality  of  the  proprietor,  and 
often  such  details  as  the  dress  of  the  sales  force  are  in  harmony 
with  a  certain  predetermined  plan. 

Among  the  points  which  should  be  given  consideration  by  a 
shop  proprietor  in  planning  a  distinctive  store,  may  be  mentioned 
the  following: 

Design  of  the  exterior;  woodwork  of  the  interior;  color  of 

*  C  I,.  Law  and  A.  J.  Marshall,  "The  Lighting  of  a  Large  Store,"  Vol.  VI  (1911), 
p.  186;  H.  W.  Shalling.— "Department  Store  Lighting,"  Vol.  VIII  (1913),  p.  17. 


LAW   AND   POWELL:     STORE   LIGHTING  51/ 

walls  and  ceiling;  finish  of  show  cases;  floor  covering;  finish 
and  type  of  lighting  fixtures,  glassware  and  lamps. 

Numerous  examples  of  distinctive  store  lighting  have  un- 
doubtedly come  to  the  attention  of  every  one,  but  a  description 
of  a  few  which  the  authors  have  observed  may  indirectly  suggest 
schemes  which  will  prove  of  benefit  and  aid  in  the  advancement 
of  the  art  of  lighting.  In  last  year's  paper,  the  quantitative 
element  in  designing  the  lighting  was  discussed.  A  number  of 
stores  were  grouped  under  one  heading,  but  with  the  case  in 
hand  where  individuality  of  stores  of  any  one  type  is  the  essen- 
tial, it  is  necessary  to  use  the  "case"  system,  illustrating  by 
example. 

TOY  STORE. 

An  effective  lighting  system  of  a  toy  store  may  be  seen  in  F. 
A.  O.  Schwartz's  store  on  Fifth  Avenue,  New  York  City.  (Fig. 
1).  The  building  is  of  modern  construction  with  a  high  ceiling 
supported  by  pillars.  The  entire  interior,  including  walls,  ceil- 
ing and  show  cases,  is  finished  in  white,  affording  an  excellent 
background  for  the  varicolored  toys  on  exhibition.  The  store 
was  formerly  lighted  by  Nernst  lamps  with  massive  ornamental 
housings  finished  in  gilt.  These  were  of  Renaissance  design, 
and  in  keeping  with  the  capitals  of  the  columns  in  the  main  room. 
The  heart-shaped  Nernst  globe  was  replaced  by  a  14  in.  (35.56 
cm.)  opalescent  glass  acorn  type  diffuser,  and  400-watt  clear 
tungsten  lamps  were  used.  Diffusion  is  good  and  shadows  from 
pillars  and  overhanging  shelves  are  minimized. 

Store  Window 

Length  145  ft.  (44.20m.).  60-watt   bowl  frosted  tungsten 

Width  41  ft.  (12.50  m.).  lamps  set  in   recessed   mirrored 

Approximate     area     6,000    square  pockets  at  front  edge  of  window, 

feet  (557.42  sq.  m.).  spaced  about  3  ft.  (0.914  m.). 

Ceiling  height  17  ft.  (5.18  m.). 
Lamps  12  ft.  (3.66  m.)  from  floor. 
15    400-watt,      4     250-watt     clear 

tungsten  lamps. 
Total  wattage  7,000. 
Watts  per  sq.  ft.  (0.30  sq.  m. )  1.2. 

JEWELRY  STORE. 
Richness  and  splendor  are  symbolized  by  jewels  and,  therefore, 
the  shop  dealing  in  these,  should  be  magnificently  finished.      The 
10 


518  TRANSACTIONS  I.  E.   S. — PART  II 

store  of  E.  M.  Gattle,  on  Fifth  Avenue,  New  York,  Fig.  2  can 
well  be  used  as  an  illustration.  All  of  the  show  cases  and  fur- 
niture are  of  mahogany;  immense  gray  marble  columns  and 
pilasters  with  gold  capitals  support  a  paneled  ceiling,  which  is 
also  of  mahogany  finish.  The  parts  of  the  side  walls  not  oc- 
cupied by  window  space  are  a  green  tint,  decorated  in  gold. 
The  floor  is  of  oak  in  parquet  style.  Light  is  furnished  by 
eighteen  shower  fixtures,  verde  finish,  using  40-watt  all  frosted 
round  bulb  tungsten-filament  lamps,  and  in  the  paneled  recesses 
in  the  front  part  of  the  store  are  eight  cut  glass  hemispheres, 
accommodating  two  25-watt  clear  tungsten  lamps  each.  A  high 
wattage  is  necessary  with  this  system  in  a  room  of  this  style,  as 
the  reflection  coefficients  of  the  ceiling  and  walls  are  very  low. 

Store  Window 

Length  78  ft.  (23.77  m.).  2  aluminum  finish    trough    refiec- 

Width  38  ft.  (11.58m.).  tors. 

Area  2,960  sq.  ft.  (274.99  sq.  m.).  25-watt   tungsten  lamps    spaced 

Ceiling  height  13  ft.  (3.96  m.).  about  10  in.  (25.4  cm.). 

Lamps  10  ft.  (3.04  m.)  from  floor.  Backing  of  window  green  plush. 

220     40-watt      round      bulb     all 

frosted  16  25-watt  clear  tungsten 

lamps. 
Total  watts  9,200. 
Watts  per  square  foot  3.1. 

TOGGERY  OR  HABERDASHERY  SHOP. 

As  this  class  of  store  caters  entirely  to  men,  the  store  fittings 
should  not  be  radical  to  any  appreciable  extent.  Neatness,  sim- 
plicity, and  up-to-date  appearance  should  characterize  the  shop. 
The  lighting  system  must  be  quite  efficient,  as  a  high  intensity 
of  illumination  is  desirable. 

Fig.  3  shows  a  night  view  of  the  installation  of  one  of  the 
shops  of  Weber  &  Heilbroner  on  Broadway,  New  York  City, 
which  conforms  excellently  with  the  above  requirements.  Six- 
arm  brush  brass  fixtures  of  well  balanced  proportions  are  used 
with  clear  100-watt  tungsten  lamps  and  opalescent  bowl  shaped 
reflectors.  Show  cases,  counters  and  woodwork  are  of  polished 
mahogany;  ceiling  smooth  white  plaster;  walls  above  shelves 
covered  with  green  burlap,  and  floor  of  hard  wood.  The  window 
trim  is  of  Circassian  walnut,  forming  an  excellent  contrast  to  the 


- 


6')* 


Fig.  i. — Distinctive  illumination  of  a  toy  store. 


Fig.  2.  — Distinctive  illumination  of  a  jewelry  store. 


Fig-.  3. — Distinctive  illumination  of  a  haberdasher's  store. 


Fig.  4. — Distinctive  illumination  of  a  millinery  store. 


LAW  AND  POWELL:     STORE)  LIGHTING  519 

dark  blue  velvet  backing  for  the  goods  on  display.  A  white  fixed 
shade,  extending  to  within  six  feet  (1.83  m.)  of  the  sidewalk 
level,  serves  as  a  valance. 

Store  Window 

Length  69  ft.  (21.03  rn.).  100-watt  clear  tungsten  lamps. 

Width  18  ft.  (5.49  m.).  Concentrating    prismatic  reflector. 

Area  1,342  sq.  ft.  (124.77  sq.  m.).  Spaced  14  in.  in  a  row  along  center 

Ceiling  height  13  ft.  (3.96  m.).  of  false  ceiling. 

Lamps  10  ft.  6  in.  (3.20  m.)  from 

floor. 
30  100-watt  clear  tungsten  lamps. 
Total  watts  3,000. 
Watts  per  square   foot  2.4. 

MILLINERY. 
Since  Paris  is  the  seat  of  fashions,  to  create  the  proper  atmos- 
phere, the  display  room  should  be  "Frenchy"  in  character. 
Mme.  Bruck's  shop  on  West  Fortieth  Street,  New  York  City, 
shown  in  Fig.  4  may  be  taken  as  an  example.  White  show 
cases,  covered  with  mirrors  line  the  walls,  and  the  dainty  furni- 
ture is  all  finished  in  white  enamel.  White  has  the  advantage 
that  it  does  not  "clash"  with  the  colored  materials  of  the  hats  and 
tend  to  divert  the  attention  from  the  goods  on  display.  The  ceil- 
ing is  of  smooth,  white  plaster  and  a  border  of  satin  finish  wall 
paper  matches  the  old  rose  Wilton  carpet  and  silk  window  hang- 
ings. Two  ten-light  brass  finish  shower  fixtures  with  bowl- 
frosted  tungsten  lamps  surrounded  with  crystal  beaded  glass, 
furnish  general  illumination.  Localized  illumination  at  the 
mirrors  is  supplied  by  side  wall  brackets,  brush  brass  finish,  Em- 
pire style,  equipped  with  bowl-frosted  tungsten  lamps,  shielded 
by  crystal  and  old  rose  beaded  shades.  A  few  plants  add  to 
the  attractiveness  of  the  room. 

Store  Window 

Length  40  ft.  (12.19  m.).  25-watt   clear   tungsten    lamps    in 

Width  12  ft.  (3.66  m.).  concentrating  prismatic  reflectors 

Area  480  sq.  ft.  (44.59  sq.  m.).  on  2  ft.  centers. 

Ceiling  height  10  ft.  (3.04  m. ).  3  25-watt  tungsten  lamps  in  crystal 

Lamps  9  ft.  (2.74  m.)  from  floor.  fixtures  in  center  of  window. 

Total  watts  750.  2  side  wall  brackets,  cut  glass  shade 

30   25-watt  bowl  frosted  tungsten  and  25-watt  tungsten  lamps. 

lamps. 
Watts  per  square  foot  1.5. 


520  TRANSACTIONS  I.  E.   S. — PART  II 

CANDY  STORES. 
One  of  the  newest  and  most  attractive  of  New  York's  Fifth 
Avenue  stores  is  that  of  Schrafft,  a  view  of  which  is  shown  in 
Fig.  5.  A  combination  of  semi-indirect  and  totally  indirect 
illumination  is  used.  The  front  portion  of  the  store  serves  as  a 
shop  and  is  lighted  ,by  five  three-light  carved  alabaster  bowls 
suspended  from  the  ceiling  by  silk-covered  supports,  and  four 
one-light  bowls,  two  on  brackets  and  two  on  short  pillars.  The 
ceiling  here  is  tan  decorated  with  raised  gold  figuring ;  walls  are 
elaborately  decorated,  with  red,  green  and  blue  on  a  neutral 
backing.  Show  cases  are  of  Circassian  walnut;  pillars  and  floor 
of  marble.  A  number  of  small  decorative  standards  are  used 
to  illuminate  the  counters.  The  rear  half  of  the  store  is  used  as 
a  lunch  room.  In  the  center  of  this  room  is  what  is  apparently 
a  fern-covered  urn.  This  contains  a  white  enamelled  reflector 
and  a  cluster  of  clear  lamps,  the  light  from  which  is  directed  to 
the  cream  colored  ceiling  and  walls,  lighting  the  room  indirectly. 

Store  Window 

Length  74  ft.  (22.55  m.).  Finished  in  Circassian  walnut;  roof 

Width  (average)  17  ft.  (5.18  m.).  recessed  with  mirrored  pyramidal 

Area  1,200  sq.  ft.  (111. 48  sq.  m.).  reflectors  and  25-watt  clear  tung- 

Ceiling  height  14  ft.  (4.27  m.).  sten  lamps  installed  in  squares  on 

Lamps  10  ft.  (3.05  m.).  18  in.  centers. 

5  150- watt  clear  tungsten  lamps. 

4  40-watt  clear  tungsten  lamps. 
15  60-watt  clear  tungsten  lamps. 
Total  watts  1,810. 
Watts  per  sq.  ft.  1.5. 

Delicious  sweets  of  great  variety  originate  in  the  Far  East, 
and  an  Oriental  scheme  of  decoration  for  a  candy  store  is,  there- 
fore, often  appropriate.  Fig.  6  shows  a  night  view  of  the  ex- 
terior of  Page  &  Shaw's  Fifth  Avenue  Shop.  It  can  be  seen  that 
the  window  is  partially  covered  with  a  delicate  tracery  of  red, 
green  and  blue  leaded  glass ;  at  night  this  is  accentuated  by  illum- 
ination from  lamps  in  the  ceiling  of  the  window.  Three  metal 
and  art  glass  hanging  fixtures  are  also  part  of  the  window  equip- 
ment. 

Free  use  of  the  primary  colors  is  made  in  the  decorating  of  the 
walls  and  ceilings  of  the  store  with  conventional  Moorish  figures. 
The  floor  is  of  composition,  red  and  white  mosaic.  It  can  be 
safely  said  that  no  two  of  the  interior  lighting  units  are  alike: 
Oriental  metal  and  colored  glass  domes,  pottery  vases  lighted 


LAW  AND  POWELL:     STORE  LIGHTING  521 

from  within  and  silk-covered  lanterns  furnish  a  very  low  inten- 
sity of  general  illumination,  with  a  higher  value  on  the  counters 
and  show  cases. 

The  cashier's  desk  is  surrounded  by  leaded  glass  made  in  the 
form  of  a  miniature  Turkish  house,  the  whole  surface  of  which 
is  illuminated  by  a  number  of  line  source  tubular  tungsten-fila- 
ment lamps  concealed  in  its  interior. 

Store  Window 

Length  30  ft.  (9.14  m.).  Roof  recessed  with  mirrored  pyra- 

Average  width  15  ft.  (4.57  in.)-  midal  reflectors  ;  one  ft.  centers  ; 

Area  450  sq.  ft.  (42.8  sq.  m.).  clear  16  c-p.  round  bulb  carbon 

Ceiling  height  10  ft.  (3.05  m.).  lamps;     2     60-watt    all     frosted 

Lamps  5  to  7  ft.  (1.52  to  2.13  m.)  round  bulb  and  3  25-watt  regu- 

from  floor.  lar   tungsten   lamps  in  hanging 

Total  watts  900.  lanterns. 
Watts  per  square  foot  2.0. 

GROCERY  STORE. 

A  neat,  attractive  display  will  cause  trade  to  flock  to  the  store 
which  is  properly  arranged.  Cleanliness  is  a  very  important 
point  to  remember.  There  is  no  demand  for  a  system  of  dec- 
oration for  this  class  of  store,  but  the  walls,  pillars  and  ceiling 
should  have  frequent  painting.  A  dark  wainscoting,  the  color 
of  the  shelves  and  show  cases,  with  neutral  walls  and  ceiling, 
makes  an  attractive  combination.  Almost  any  lighting  unit  which 
is  neat  and  inconspicuous  will  serve. 

D.  M.  Welch  &  Son's  store  in  New  Haven,  Conn.,  shown  in 
Fig.  7  serves  as  an  illustration  of  the  above  requirements.  The 
counters  and  show  cases  are  of  hard  wood,  natural  finish ;  the 
trim  is  dark  green ;  and  the  ceiling  and  walls  are  painted  a  light 
tint.  Neatness  is  particularly  characteristic.  Two  hundred 
and  fifty  watt  tungsten  lamps,  in  totally  enclosing  prismatic  re- 
flectors are  used  for  general  illumination.  The  unit  is  efficient 
and  a  satin  finished  lower  half  provides  excellent  diffusion.  A 
short  brush  brass  chain  with  canopy  serves  as  the  fixture. 

Store  Window 

Length  70  ft.  (21.33  m.).  100-watt  clear  tungsten   lamps   in 

Width  40  ft.   (12.19  m.).  concentrating    prismatic    reflec- 

Area  2,800  sq.ft.  (232.25sq.m.).  tors  spaced  2  ft.  6  in.  (0.76  m.). 

Ceiling  height  14  ft.  (4.27  m.). 
Lamps    11    ft.    (3.25    m.)    from 
floor. 

10  250-watt  tungsten  lamps. 

Total  watts  2,500. 

Watts  per  square  foot  0.9. 


522  TRANSACTIONS  I.  E.  S. — PART  II 

TEA  ROOM. 

Coziness  is  the  keynote  of  success  of  these  establishments.  A 
number  of  years  ago  a  young  woman  started  in  a  small  way  to 
sell  home-made  candy  and  pastry  among  her  friends.  Her 
energies  soon  developed  into  a  methodical  business  system,  and 
her  products  sprang  rapidly  into  favor,  with  the  result  that 
"Mary  Elizabeth"  has  branch  shops  in  many  of  the  large  cities. 
Her  New  York  store,  which  is  shown  in  Fig.  8  and  located  on 
Fifth  Avenue,  is  finished  in  white  on  the  outside,  with  her  fac- 
simile signature  in  black  serving  as  a  sign. 

The  shop  itself  is  modelled  after  a  New  England  interior  of 
fifty  years  ago;  the  ceiling  is  low  and  finished  in  white  plaster; 
the  floor  of  wide  boards  is  painted  a  dark  yellow  and  covered 
here  and  there  with  rag  carpet  "runners".  The  tea  room  proper 
is  in  the  rear.  On  the  right  is  an  old  fashioned  fire  place,  and 
on  the  left  a  number  of  "stalls"  similar  to  those  found  in  taverns 
of  bygone  days.  Small  tables,  covered  with  spotless  linen,  and 
gilt  chairs  are  arranged  as  shown  in  the  illustration.  Shelves, 
counters  and  windows  are  trimmed  with  dainty  white  material. 

Light  is  furnished  by  tungsten  lamps  in  shirred  silk  shades 
which  have  a  slight  touch  of  color.  Sixteen  of  these  are  at- 
tached to  ceiling  outlets  and  eight  are  on  wall  brackets. 

The  atmosphere  of  the  room  is  extremely  inviting  and  the 
scheme  of  decoration  well  executed. 

Store  Window 

Length  62  ft.  (18.90  m.).  40-watt  tungsten  lamps  in  shades 

Width  (average)  18  ft.  (5.49m.).  as  used  in  the  store;  row  in  the 

Area  1,110  sq.  ft.  (103. 11  sq.  m.).  center  of  ceiling;    2  ft.  (0.61  m.) 

Ceiling  height  8  ft.  (2.44  m.).  centers. 

Lamps    7    ft.     (2.13  m.)     from 

floor. 
24  40-watt  clear  tungsten  lamps. 
Total  watts  960. 
Watts  per  square  foot  0.9. 

SHOE  STORE. 
Most  stores  of  this  class  have  a  center  bench  arrangement,  the 
entire  wall  space  being  covered  with  boxes  on  shelves.  A  room 
of  medium  width  will  require  at  least  two  rows  of  units  to  give 
satisfactory  illumination  on  the  labels  on  the  boxes  and  at  the 
foot  rests  where  the  shoes  are  fitted  and  inspected. 

A  particularly  novel  layout  is  shown  in  Fig.  9,  a  night  view 


f>' 


Fig.  5. — Distinctive  illumination  of  a  candy  store. 


Fig.  6. — Distinctive  illumination  of  a  candy  store  window. 


Fig.  7. — Distinctive  illumination  of  a  grocery  store. 


Fig.  8.— Distinctive  illumination  of  a  tea  room. 


LAW  AND  POWELL:     STORE  LIGHTING  523 

of  Frank  Brothers'  Fifth  Avenue  (New  York)  shop.  Entering 
from  the  street,  one  passes  into  the  rotunda  (shown  in  the  back- 
ground of  the  photograph)  about  16  feet  in  diameter,  the  dome 
of  which  is  supported  by  Corinthian  columns.  The  floor  is  of 
mosaic  marble  and  the  ceilings,  cream  colored,  with  raised  plaster 
decorations.  Show  cases,  with  attractive  dressings,  are  grouped 
about  the  room.  Suspended  from  the  center  of  the  dome  is  an 
ornamental  inverted  fixture  containing  eighteen  lamps.  This 
consists  of  six  diffusing  glass  globes,  pressed  into  the  form  of 
huge  shells;  below  these  are  four  round  bulb  carbon  lamps  en- 
closed in  amber  beaded  glass. 

The  store  proper  is  rectangular  in  shape  and  a  balcony  6  feet 
(1.83  in.)  wide  extends  completely  around  the  interior.  The 
cream  colored  ceiling  beneath  the  balcony  is  divided  by  beams 
into  squares.  In  the  center  of  each  square  is  a  lighting  fixture 
consisting  of  five  pieces  of  pearl-like  glass  in  the  form  of  a  large 
shell ;  a  40-watt  clear  tungsten  lamp  is  located  above  each  shell. 
At  the  base  of  the  shell  is  a  25-watt  round  bulb  all  frosted 
tungsten  lamp. 

On  each  pillar  from  the  balcony  to  the  ceiling  are  located  two 
two-arm  brass  brackets  with  clear  gem  lamps  in  roughed 
glass  spheres.  These  serve  to  light  the  balcony  and  the  center 
portion  of  the  store  proper. 

The  oak  parquet  floor  is  partly  covered  with  rugs :  the  furni- 
ture is  leather  covered  and  the  showcases  and  shelves  are  of 
mahogany. 

store  The  value  of  watts  per  square 

Length  64  ft.  (19.5  m.).  foot   would   be   of   little   sig. 

Width  24  ft.  (7.31  m.).  nicance,  as  two  types  of  lamps 

Area   (main  floor)    1,540  sq.  ft.                    are  in  use>  and  also  both  the 

(14307  sq.  m.).  balcony    and    main    floor    are 

Height     under     balcony  8    ft.                   lighted 

2.44  m.). 

Height     above    balcony     10    ft. 

(3-05  m.). 
17  25-watt  round  bulb  tungsten 

lamPS-  Windows 

85  40-watt  clear  tungsten  lamps.  Mirrored    trough     reflector    with 

64  50-watt  clear  Gem  lamps.  50-watt   Gem   lamps   outlets   on 

Total  watts  7,000.  9  in.  centers. 


524  TRANSACTIONS  I.  %.   S. PART  II 

ANTIQUE  AND  CURIO  SHOP. 

In  many  cases  the  lighting  requirements  of  an  antique  store  are 
similar  to  those  for  a  high  class  furniture  store,  that  is,  a  low  in- 
tensity of  diffused  light  suffices.  Exposed  light  sources  are  very 
obj.ectional,  as  the  polished  surfaces  show  the  reflection  and  glare 
is  to  be  deplored  in  viewing  the  rare  pieces  on  exhibition.  Quite 
often  the  lighting  units  themselves  are  "objects  d'art."  Such  is 
the  case  in  the  shop  of  Lewis  &  Simmons,  shown  in  Fig.  10, 
where  hand  carved  alabaster  bowls  with  clusters  of  clear  lamps 
furnish  semi-indirect  illumination. 

The  white  ceiling,  walls  covered  with  dark  red  velvet  and  tan 
velvet  carpet,  make  a  good  color  combination  for  displaying  the 
goods  by  contrast. 

Store  Window 

Length  48  ft.   (14.63  m.).  Mirrored  trough  reflector  with  25- 

Width  (average)  13  ft  (3.96  m).  watt  clear  tungsten  lamps  spaced 

Area  630  sq.  ft.   (58.57  sq.  m.).  8  in.   (0.20  m.)    apart  along  the 

Ceiling  height  14  ft.   (4.27  m.).  top  of  windows,  and  upright  at 

Lamps    8    ft.     (2.44    m.)     from  the    two    sides    to    a    height    of 

floor.  about  4  ft.  (1.22  m.). 
20  40-watt  clear  tungsten  lamps. 
Total  watts  800. 
Watts  per  square  foot  1.3. 

BOOK  STORE. 
Scribners'  new  store,  Fifth  Avenue,  New  York,  is  an  excellent 
demonstration  of  a  carefully  planned  and  well-executed  scheme 
of  lighting.  The  illustration,  Fig.  11,  shows  very  well  the  general 
appearance  of  the  room.  The  ceiling  of  the  main  bay  is  vaulted 
and  is  of  light  gray  sandstone  with  white  plaster  panels.  This 
is  lighted  by  means  of  line  source  tungsten  lamps  (approximately 
25  watts  per  foot)  the  reflectors  being  located  above  the  mould- 
ing running  around  the  cove.  Fourteen  opalescent  glass  bowls; 
equipped  with  clusters  of  three  lamps  each,  hung  from  the  ceiling 
by  long  brass  rods,  furnish  a  feature  which  seems  desirable,  vis., 
a  visible  source  of  illumination.  The  book  racks  and  balconies 
in  the  side  bays  are  lighted  by  60-watt  clear  tungsten  lamps  in 
opalescent  bowl-shaped  reflectors.  Paintings  on  the  rear  wall 
are  lighted  by  individual  mirrored  trough  reflectors  equipped  with 
25-watt  clear  tungsten  lamps  on  one  foot  centers.     The  entire 


Fig.  9. — Distinctive  illumination  of  a  shoe  store. 


Fig.  10.— Distinctive  illumination  of  an  antique  store. 


Fig.  ii. — Distinctive  illumination  of  a  book  store. 


pig.  12.— Distinctive  illumination  of  a  ladies-  wear  store. 


LAW   AND   POWELL:     STORE   LIGHTING 


525 


front  of  the  store  is  of  plate  glass,  thus  furnishing  an  excellent 
supply  of  daylight,  and  the  cases  and  shelves  being  of  light  oak 
give  the  room  a  most  cheerful  appearance. 

Windows 


Store  (main  bay) 
Length  98  ft.   (29.87  m.). 
Width  28  ft.    (8.53  m.). 
Area  2,740  sq.  ft.  (254.5  sq.  m.). 
Ceiling    height    (maximum)    30 

ft.  (9.14  m.). 
Lamps    9    ft.     (2.74    m.)     from 

floor. 
128  35- watt  tubular  tungsten 

lamps. 
42  40-watt  clear  tungsten  lamps. 
Total  watts   (approx.)   7,580. 
Watts  per  square  foot  2.8. 


No  special  lighting,  as  the  win- 
dows extend  to  the  top  of  the 
arch  and  the  whole  store  is  a 
flood  of  light. 


LADIES'  WEAR. 

This  type  of  shop  is  really  divisible  into  two  classes:  namely, 
general  and  specialized. 

As  an  example  of  the  first  class,  the  Fifth  Avenue  (New  York) 
store  of  J.  M.  Gidding,  which  is  shown  in  Fig.  12,  may  be  given  a 
little  attention.  The  lighting  units  are  of  the  sunburst"  type, 
consisting  of  6  regular  and  6  round  bulb,  all-frosted  tungsten 
lamps  below  a  gilded  composition  plate,  all  suspended  by  a  single 
chain.  The  fixtures  are  pleasingly  harmonious  with  the  cream 
colored  ceiling  and  delicate  gold  lining.  Show  cases  and  wood- 
work are  of  magnificent  Circassian  walnut,  which  blends  well 
with  the  rich  carpet  of  green  and  tan.  The  wall  visible  above 
the  dressing  room  is  largely  covered  with  gilt  figures. 

Window 

Mirrored    trough     reflector    with 
25-watt  tungsten  lamps  on  9  in. 


Store 
Length  55  ft.  (16.76  m.). 
Width  48  ft.  (14.63  m.). 
Area  2,640  sq.  ft.  (245.25  sq.  m.). 
Ceiling  height  12  ft.  (3.66  m.). 
Lamps    9    ft.     (9.74    m.)     from 

floor. 
90  20-watt   all-frosted   tungsten 

lamps. 
90  25-watt   all-frosted   tungsten 

lamps. 
Total  watts  4,050. 
Watts  per  square  foot  1.5. 


centers.  White  false  ceiling, 
with  two  alabaster  carved  bowls, 
equipped  with  6  40-watt  tungsten 
lamps  each,  are  suspended  from 
this.  Base  and  trim  of  window 
Circassian   walnut. 


526  TRANSACTIONS  I.  £.   S. PART  II 

The  specialized  ladies'  wear  shop  is  exemplified  by  the  shop 
of  W.  B.  Crocker,  Fig.  13,  which  handles  mourning  goods  ex- 
clusively. The  scheme  of  decoration  is  very  appropriate.  As  one 
observer  remarked,  "A  sombre  cheerfulness  fills  the  room."  A 
rich  gray  carpet  is  but  a  shade  darker  than  the  wall  covering, 
which,  in  turn,  matches  the  woodwork  of  the  show  cases  and  trim. 
The  chairs  are  of  gray  oak  and  tables  of  wicker  work. 
The  ceiling  is  white,  and  suspended  from  this  by  long  chains  are 
four  seven-light  and  one  fourteen-light  shower  fixtures,  dull  sil- 
ver finished.  Low  wattage,  bowl-frosted  tungsten  lamps  are  used 
with  diffusing  shades. 

Under  the  balcony  at  the  rear  of  the  store  are  full-length  mir- 
rors. Localized  illumination  is  provided  at  each  by  a  25-watt 
all-frosted  round  bulb  tungsten  lamp.  In  the  front  portion  of  the 
store,  the  general  illumination  is  supplemented  by  two-arm 
brackets  similar  in  finish  and  equipment  to  the  overhead  units. 

Store  Window 

Length  76  ft.   (23.16  m.) .  Mirrored      trough      reflectors 

Width  16  ft.   (4.88  m.).  equipped     with     50-watt     Gem 

Area  1,210  sq.  ft.  (112.4  sq.  m.).  lamps  on  9  in.  centers.     Wood- 

Ceiling  height  18  ft.   (5.48  m.).  work    gray.      Mirrors    at    side. 

Lamps    10   ft.    (3.05    m.)    from  Beaded    crystal    hemisphere    set 

floor,  in  center  of  the  ceiling. 

60  15-watt  bowl-frosted  tungsten 

lamps. 
18  25-watt  round  bulb  tungsten 

lamps. 
Total  watts  1,390. 
Watts  per  square  foot  1.15. 

STATIONERY. 
When  mention  is  made  of  this  class  of  store,  one  involuntarily 
pictures  in  his  mind  a  low-ceilinged,  dingy  room  with  everything 
arranged  in  a  haphazard  manner;  cigars,  newspapers,  candy  and 
stationery  in  a  grand  mix-up.  In  contrast  to  this,  is  is  pleasing  to 
note  Dennison's  new  store  on  Fifth  Avenue  (New  York),  Fig.  14. 
Immense  square  columns  support  a  pure  white  ceiling  beautifully 
decorated  with  raised  plaster  figures.  The  woodwork  of  shelves, 
drawers,  show-cases  and  counters  is  of  carefully  selected 
weathered  oak.  Neatness  is  the  predominating  feature,  and  the 
semi-indirect  lighting  units  of  canary  and  white  glass  harmonize 


5^ 


Fig.  13. — Distinctive  illumination  of  a  ladies'  wear  store. 


Fig.  14.— Distinctive  illumination  of  a  stationery  store. 


Fig.  15. — Distinctive  illumination  of  a  restaurant. 


Fig.  16. — Distinctive  illumination  of  delicacy  store. 


LAW  AND  POWELL:     STORE  LIGHTING  527 

perfectly  with  this  characteristic.  The  single-chain  suspension 
and  the  bowl  itself  are  designed  along  lines  of  simplicity.  Six 
lamps  are  installed  in  each  fixture. 

A  balcony  is  seen  at  the  rear  of  the  store ;  this  is  used  for 
office  purposes  and  is  lighted  by  four  four-light  units  with  short 
ceiling  suspension  similar  in  design  to  the  large  units. 

The  space  below  the  balcony  has  been  given  quite  a  bower-like 
appearance  by  the  use  of  a  false  ceiling  of  green  lattice  work  and 
a  profusion  of  paper  flowers. 

Store  Window 

Length  67  ft.   (20.42  m.).  Mirrored  trough  reflector  equipped 

Width  36  ft.   (10.97  m.).  with    tungsten    lamps    25    watts 

Total  area  2,410  sq.  ft.    (223.88  foQt 

sq.  m.). 
Ceiling  height     18  ft.  (5.48  m.). 
Lamps    12    ft.    (3.66   m.)    from 

floor. 
60  100-watt  clear  tungsten  lamps. 
Total  watts  6,coo. 
Watts  per  square  foot  2.5. 

RESTAURANT. 
There  seems  to  be  no  definite  practise  with  regard  to  the  light- 
ing or  rooms  of  this  nature.  Some  proprietors  desire  a  great 
flood  of  light  and  the  attendant  sparkle  as  produced  by  crystal 
chandeliers;  others  demand  a  soft,  well-diffused,  low  general 
illumination  supplemented  by  localized  table  lamps.  Bergfield's 
restaurant,  on  Broadway,  New  York,  Fig.  15,  is  an  example  of 
lighting  with  the  latter  idea  of  proper  lighting.  Totally  indirect 
single  unit,  mirrored  reflector  lighting  units  of  composition 
moulded  into  an  Egyptian  design  are  used.  Mirrors  are  set  into 
the  wall  panels,  and  above  each  is  a  two-arm  verde  finished 
bracket  with  low-wattage  multiple  lamps  and  silk  shades.  The 
wood  work  is  cream,  with  gold  decorations;  ceiling  white;  wall 
panels  old  rose,  satin  finish;  chairs  mahogany  finish,  and  the 
carpet  a  neutral  green. 

Store  Store 

Length  77  ft.   (23.47  m-)-  5  250-watt  clear  tungsten  lamps. 

Width  27  ft.   (8.23  m.).  4  100-watt  clear  tungsten  lamps. 

Area  2,080  sq.  ft.  (193.23  sq.  m.).  68  10-watt  clear  tungsten  lamps. 

Ceiling  height  13  ft.   (3.97  m.).  Total  watts  2,330. 

Lamps    9    ft.     (2.74  m.)     from  Watts  per  square  foot  I.I. 
floor. 


528  TRANSACTIONS  I.  E.   S. — PART  II 

DELICACIES. 

The  Specialty  Shop  in  Boston,  Fig.  16,  has  a  dark  wainscoat- 
ing  about  6  feet  (1.83  m.)  high;  above  this  the  walls  are  divided 
into  panels ;  in  each  of  these  panels  is  set  a  heraldic  design,  which 
has  been  adopted  by  the  proprietor  as  a  trade  mark.  The  ceiling 
is  white,  glazed,  and  divided  into  polygons  by  the  moulding.  The 
show  cases  and  counters  are  divided  into  panels  which  are  prac- 
tically replicas  of  the  wall  panels.  The  floor  is  mosaic  tile.  The 
lighting  system  is  remarkably  in  accord  with  the  general  scheme. 
Between  adjacent  panels  is  a  torch  fixture  with  an  upright  lamp 
and  diffusing  ball.  Counter  standards,  wall  brackets  and  four- 
arm  ceiling  fixtures  carry  bowl-frosted  tungsten  lamps  and  pyra- 
midal art  glass  shades  which  are  finished  to  match  the  counter 
and  wall  trim.     A  most  pleasing  harmony  is  secured. 

As  proof  that  original  store  layouts  are  noticed  by  the  general 
public,  the  management  of  this  store  reports  that  the  lighting 
is  the  subject  of  many  favorable  remarks,  both  by  local  and 
out-of-town  customers. 

Store. 
Length  72  ft.  (21.94  m-)- 
Width  40  ft.   (12.19  m.). 
Area  2,880  sq.  ft.  (267.55  sq.  m.). 
Ceiling  height  20  ft.    (6.09  m.). 
Lamps    t8    ft.    (5.48    m.)    from 

floor. 
25  60-watt  tungsten  lamps. 
14  25-watt  tungsten  lamps. 
Total  watts  1,850. 
Watts  per  square  foot  0.65. 

CONCLUSION. 

A  sufficient  number  of  individual  installations  have  been  de- 
scribed to  indicate  quite  clearly  that  the  lighting  system  should 
blend  with  the  general  scheme  of  decoration.  The  illuminating 
equipment,  rather  than  being  dazzling,  glaring  or  commonplace, 
is  inconspicuous,  and  forms  a  part  of  the  furnishing  of  the  room. 

It  must  be  borne  in  mind  that  the  methods  outlined  above  are 
not  the  only  correct  schemes  of  lighting  to  use;  often  in  the  lay- 
ing out  of  an  installation  the  ideas  or  desires  of  the  proprietor 


LAW  AND  POWELL:     STORE  LIGHTING  529 

will  produce  considerable  deviation  from  the  scheme  which  would 
be  most  in  keeping  with  the  period  of  architecture  that  is  being 
followed. 

From  the  descriptions  given,  it  can  be  seen  that  one  is  able  to 
apply  the  commercial  diffusers  and  reflecting  devices  to  almost 
any  class  of  service.  As  far  as  possible  the  endeavor  has  been 
made  to  discuss  stores  which  had  standard  equipment,  thus  show- 
ing that  there  is  no  necessity  for  the  design  of  special  auxiliaries. 
Expanding  this  idea  slightly,  the  authors  believe  that  the  stores 
described  are  distinctive  and  yet,  with  the  exception  of  the  carved 
alabaster  bowls,  the  initial  cost  is  relatively  low. 

These  illustrations  might  have  been  continued  at  great  length 
and  an  appropriate  use  found  for  almost  all  the  equipment  listed, 
but  this  is  obviously  out  of  the  question,  so  the  paper  can  well  be 
closed  with  the  admonition. — In  designing  the  lighting  for  shops 
of  the  class  treated  in  this  paper,  use  discretion  in  the  selection 
of  lighting  units  and  do  not  offer  the  prospective  customer  some- 
thing which  is,  on  the  face  of  it,  purely  utilitarian. 

The  authors  desire  to  thank  the  photographic  bureau  of  the 
New  York  Edison  Company,  for  their  assistance  in  taking  pho- 
tographs and  making  autochromes  and  lantern  slides. 

DISCUSSION. 

Mr.  M.  H.  Flexner:  Knowing  that  the  larger  units  are 
more  efficient  than  the  smaller  ones;  that  just  as  good  results 
can  be  accomplished  with  the  larger  units — and  I  am  satisfied 
that  equal  artistic  effects  can  be  obtained  as  with  the  smaller 
ones — I  would  like  to  ask  Mr.  Powell  why  the  clusters  seem 
to  be  so  much  in  evidence? 

Mr.  S.  G.  Hibben  :  It  seems  from  the  foregoing  paper  that 
most  of  these  distinctively  lighted  stores  have  had  their  lighting 
fixtures  built  up  to  be  in  harmony  with  the  interior  decoration. 
It  would  be  excellent  to  have  fixtures  and  surrounding  decora- 
tions planned  and  built  up  simultaneously,  as  is  now  being  done 
by  some  of  the  large  department  stores.  This  brings  forth  the 
advantages  of  co-operation  between  the  lighting  engineer,  and 
the  architect,  or  particularly  the  interior  decorator. 


530  TRANSACTIONS  I.  E.   S. PART  II 

I  believe  that  of  the  five  or  six  problems  of  residence  light- 
ing that  I  come  in  touch  with  every  day,  there  are  only  perhaps 
one  or  two  that  are  new  installations,  and  it  seems  to  me  that 
the  illuminating  engineer  is  called  on  for  advice  only  when  the 
room  or  building  is  so  poorly  lighted  that  something  must  be 
done.  He  is  a  sort  of  "lighting  doctor,"  giving  a  cure  rather 
than  a  preventative. 

It  may  be  of  interest  to  notice  how  many  types  of  glass 
appliances  are  regularly  available  for  distinctive  store  lighting. 
Quite  often  the  consumer  may  make  large  expenditures  for  a 
peculiar  or  special  design  of  glassware,  that  might  be  saved 
him,  were  he  more  thoroughly  acquainted  with  the  large  variety 
of  illuminating  glassware  on  the  market.  In  briefly  mention- 
ing some  of  these  available  types,  I  would  call  attention  to  the 
rapidly  increasing  number  and  variety  of  glass  bowls  and  semi- 
indirect  reflectors,  with  open  tops,  or  covered  with  crystal  glass 
plates,  or  partially  closed  as  in  urn  shapes. 

The  decorator  can  choose  from  a  large  variety  of  period  de- 
signs, Gothic,  Elizabethan,  Doric,  Adam,  Georgian,  William 
Morris,  Colonial,  etc.  One  can  have  glass  cylinders,  columns 
with  bases  and  capitols,  or  glass  troughs  for  outline  and  cornice 
lighting.  Flat  or  configurated  diffusing  glass  plates  are  avail- 
able for  ceiling  panels,  and  plaques  for  side  walls  that  may  serve 
to  replace  open  bracket  lights. 

The  painting  or  color  decorating  of  glass  is  another  feature 
being  developed  to  bring  out  a  relief  design  by  shading,  or  to 
give  special  monograms  in  glass  for  fraternal  orders,  clubs  or 
stores  that  feature  a  trade  design  or  coat-of-arms.  In  distinctive 
stores,  like  these  described,  the  gold  or  silver  fixtures,  or  ones  , 
with  such  finishes  may  be  matched  by  properly  colored  glass- 
ware. 

Mr.  R.  B.  Ely:  A  paper  on  distinctive  store  lighting  I  think 
should  be  encouraged.  In  some  of  the  illustrations,  particularly 
the  restaurant  where  indirect  lighting  was  employed.  I  notice 
that  some  direct  lighting  brackets  were  used.  These  brackets, 
I  think,  would  be  a  distracting  feature.  In  another  of  the 
installations,  I  notice  that  one  lamp  has  been  exposed;  this  is 
a  particularly  bad  feature  in  the  system.     Take  the  installation 


STORE  LIGHTING  531 

in  the  tea  room  with  the  silk  shades  and  the  wiring,  which  I 
believe  was  exposed.  This  artistic  installation  must  have  been 
comparatively  inexpensive.  I  also  notice  a  lack  of  portable 
lamps  in  the  installations  shown.  Such  lamps  I  think  are  com- 
ing into  more  general  use.  The  greatest  drawback  to  their  use 
is  the  care  of  the  extension  cords.  However,  in  some  instances 
where  lamps  of  this  kind  have  been  equipped  with  leaded  glass, 
and  other  artistic  designs,  they  have  been  very  effective. 

Mr.  A.  L.  Powell:  Referring  to  Mr.  Flexner's  query  as  to 
the  cause  of  the  clusters  being  used  to  such  a  great  extent  in 
the  examples  shown,  I  would  say  that  the  majority  of  the 
installations  were  designed  by  the  architect.  My  experience 
has  been  that  the  architect  prefers  the  clusters  to  the  individual 
lamp  for  the  additional  flexibility  possible  with  it,  and  also  for 
the  reason  that  if  one  particular  lamp  burns  out,  the  illumina- 
tion in  a  given  section  will  not  be  materially  decreased. 

The  advisability  of  combining  direct  and  totally  indirect  light- 
ing in  the  restaurant  was  questioned.  It  seems  that  the  use  of 
brackets  in  this  particular  case,  is  very  feasible,  for  a  relatively 
large  percentage  of  the  patrons  are  ladies,  and  they  prefer  light 
coming  from  the  side,  for  the  illumination  of  their  faces.  In 
the  case  in  hand,  rather  dense,  shirred  silk  shades  were  used, 
and  the  glare  was  absolutely  unnoticeable ;  in  fact  these  units 
added  considerably  to  the  pleasantness  of  the  room. 

The  use  of  portable  lamps  was  suggested  as  an  advisable 
feature  of  distinctive  store  lighting.  Careful  reference  will 
show  these  in  use  in  Page  &  Shaw's  and  the  Boston  Specialty 
Shop,  and  in  Schraffts  the  counter  lamps  perform  the  same 
function. 

Mr.  Hibben's  remarks,  as  to  the  use  of  tinted  opalescent 
glasses,  is  very  timely  and  there  are  many  cases  in  which  they 
fit  in  excellently  with  the  general  architectural  schemes,  and 
beautiful  effects  may  be  produced  with  them. 

The  lack  of  halation  of  the  autochrome  plate  may  be  readily 
explained.  The  ordinary  plate  is  used  with  the  emulsion  side 
toward  the  lens;  light  passes  through  the  emulsion,  strikes  the 
glass  plate,  and  being  reflected,  re-enters  the  emulsion,  pro- 
ducing halation.     The  autochrome  plate  is  placed  in  the  camera 


532  TRANSACTIONS  I.  £.   S. — PART  II 

with  the  glass  side  toward  the  lens  and  a  piece  of  dull  black 
cardboard  against  the  emulsion;  light  passes  through  the  plate, 
and  the  emulsion  and  strikes  the  dull  black  surface  of  the  card 
which  reflects  but  very  little  light,  thus  reducing  halation. 


HARRISON  AND  EDWARDS  \     INCANDESCENT  LAMPS  533 

RECENT  IMPROVEMENTS  IN  INCANDESCENT  LAMP 
MANUFACTURE.* 


BY   WARD    HARRISON   AND   EVAN   J.   EDWARDS. 


Synopsis:  i.  Increased  mechanical  strength  of  tungsten  filaments: 
The  strength  of  tungsten  filament  has  more  than  increased  300  per  cent, 
since  1908,  and  the  strength  of  drawn  wire  has  increased  40  per  cent, 
since  191 1.  Greater  strength  permits  operation  at  increased  efficiencies 
at  no  decrease  in  total  life.  2.  Better  candle-power  maintenance:  The  use 
of  chemical  in  the  bulbs  which  has  become  general  during  the  past  year 
has  reduced  the  blacking  of  lamps  to  a  marked  degree  and  it  is  therefore 
possible  to  operate  them  at  efficiencies  correct  from  the  standpoint  of 
total  life  with  no  shortening  of  the  useful  life.  The  performance  of 
chemical  lamps  is  not  satisfactory  when  operated  at  low  efficiencies. 
3.  Decreased  bulb  size:  Use  of  chemical  has  made  possible  a  substantial 
reduction  in  bulb  size  for  several  lamps.  Decreased  bulb  size  reduces 
manufacturing  costs  and  broadens  the  application  of  the  lamp.  4.  Stand- 
ardization :  During  the  past  year  lamp  dimensions  have  been  standardized 
in  every  particular.  The  average  deviation  from  the  standard  is  less  than 
one-fourth  that  of  a  year  ago.  5.  Helical  filaments :  The  introduction  of 
the  coiled  filament  makes  possible  many  new  forms  of  lamps  which  here- 
tofore could  not  be  manufactured.  The  strength  of  the  filaments  is 
increased  by  this  process  and  the  candle-power  maintenance  is  not  affected. 
The  operation  of  helical  filament  lamps  at  high  efficiencies  and  their  use 
in  small  bulbs  rather  than  poorer  performance  are  the  causes  of  their 
comparatively  low  life  ratings.  The  new  tubular  lamp  and  the  focus  type 
lamp  have  many  applications  such  as  showcase  lighting,  use  in  projec- 
tors, stereopticons  and  the  like. 

The  purpose  of  this  paper  is  to  review,  briefly,  recent  improve- 
ments in  the  art  of  incandescent  lamp  manufacture  and  their 
commercial  applications.  The  more  important  of  these  improve- 
ments may  be  grouped  under  five  heads : 

(1)  Increased  mechanical  strength;  (2)  Better  candle-power 
maintenance,  obtained  by  the  use  of  chemical  in  the  bulb;  (3) 
Decreased  bulb  size;  (4)  Standardization  of  lamp  dimensions; 
(5)  Production  of  filaments  in  helical  form. 

*  A.  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society.  Pittsburgh,  Pa.,  September  22-26,  1913. 

The   Illuminating   Engineering   Society   is  not   responsible   for  the  statements  or 
opinions  advanced  by  contributors. 

II 


534 


TRANSACTIONS  I.   E.   S. — PART  II 


STRENGTH. 
The  earliest  tungsten-filament  lamps  were  so  fragile  that, 
barring  those  infrequent  cases  in  which  the  lamps  blackened  after 
the  first  few  hours  of  service,  their  performance  was  judged 
almost  entirely  on  the  basis  of  total  life  figures.  It  is  well  known 
to  the  members  of  this  Society  that  the  strength  of  the  pressed- 
filament  lamps  steadily  increased  and  with  the  introduction  of 
the  drawn-wire  lamp  in  191 1  a  very  marked  improvement  took 
place.  Perhaps  some  do  not  realize,  however,  that  the  increased 
strength  of  drawn  wire  which  has  been  effected  during  the  past 
two  years  is  even  greater  than  the  difference  in  strength  between 
the  pressed-filament  and  the  drawn-wire  filament  of  191 1.  That 
this  is  actually  the  case  is  shown  graphically  in  Fig.  I  which  gives 


Z090 

^£60 
So 
§fc50 
5" 

«z40 

>i30 

5^20 
^°  10 

1907        1908        1909         1910         1911          1912        1913         1914 
YEAR 

the  result  of  transverse  tests  of  filaments  manufactured  in  each 
year  since  1908.  The  ordinates  of  this  curve  are  proportional  to 
the  distance  through  which  a  filament  of  given  dimensions  will 
bend  before  breaking  when  stressed  by  a  gradually  increasing 
load.  An  increase  in  the  strength  of  a  lamp  also  implies  a  more 
homogeneous  and  uniform  filament  and  the  practical  result  is  that 
the  lamps  may  be  operated  at  a  far  higher  efficiency  than  before, 
with  no  decrease  in  total  life. 

CANDLE-POWER  MAINTENANCE. 

The  other  factor  most  important  in  determining  the  useful  life 

of  an  incandescent  lamp  is  the  decrease  in  candle-power  with 

age  which  takes  place  as  the  result  of  the  blackening  of  the  bulb 

by  particles  thrown  off  by  the  filament.     While  steady  improve- 


<K 


6^ 


120 

110 
100 

BE 

£90 
£80 

370 

zfeO 

°50 

g« 

o:  JO 
u 

°-  20 

10 

0 

-> 

^ 

^ 

.... 

-2[ 

- 

-- 

-— . 

~"""*&a 

L* 

*B.a 

B.Q. 

H0UR6  LIFE 

tL-*fcJU*J 


Fig.  3- 


*.See  lines  12  to  21  011  page  547. 


«  BOOK  .CHED.  T-l 

S3  Mazda 

e Straight  Side 

le  Large 

-"■60 

™ 100-130 

=  1.12 
»   S-21-E2 

e   #101 

UCTION  FACTOR      ^rf 

.  OVERALL  LENGTH       X4O     mil     5£" 

=ENTER  CFROM  BABE)       g?     jm     3-7/I6" 

:enter  ip,o»  ,ul,  .no. 1  43nmi.il/16nl 
•  top  tube   5*  or  6  mm  e 

*TH  TOP  TUBE     §9     BOB     ft " 

'»--»'».  3  ™  (2f-3i)         7 

.  O.AM.  TOP  BUTTON    5*.  mm     UM"'     , 

D,«.  BOTTOM  BUTTON  Q     j^      20/64"^? 
TOP  SUPPORT.   6  '  ^ 

SUPPORT   DIAM.      4^      mils 

support  hook  4^-32-7-3  (closed) 

BOTTOM   SUPPORTS        ^ 


OM  SUPPORT  DIAM      I3     mils 

OM  SUPPORT  HOOK    1J_26_8_3       (olCSed) 

mum  spread    29  mm  72/64  " 
"""""    6   mm   (4-8) 

>ING     B0#      1 

™     No.    14 

er  position    All  bottom  hooks 

I  .  BTEMTUBINO      H     mm      (lO^-lli^ 

"""   19  mm  f 

STEM   LENGTH     2J      TJVm      1" 
KNESS   STEM   SEAL    Ji     mm 

'  507-I6-4-I656   Copper    Tube 

BULB  NECK  jj      «      mm     I-5/16" 
T  FORMULA      fJO.2 
.USTSCHEO.     HO.1      (5-5-2) 

o     Sohed.    Ho.    2 

PKG.OUAN.      100 
PKS.WE.GHT       33      lb8. 

PPE"  A2145 

R  CONTAINER  ^219  5 
:R  CONTAINER     M2142Q 


52-22M-60 


Sheet   Bo. 153 


Fig.  4. 


B. 

Fig-  5- 


HARRISON  AND  EDWARDS  :    INCANDESCENT  LAMPS  535 

ments  have  been  made  since  1908  in  the  direction  of  diminishing 
this  black  deposit,  except  in  the  smaller  sizes,  they  have  in  no 
way  kept  pace  with  the  increasing  strength  and  uniformity  of 
the  lamp  filaments  and,  owing  to  this  blackening  of  the  bulb, 
which  becomes  greatly  accelerated  at  high  filament  temperatures, 
it  has  been  found  impossible  to  operate  many  of  the  lamps  at  an 
efficiency  warranted  by  their  total  life  performance.  To  be  of 
practical  value,  therefore,  further  improvements  in  the  quality 
of  incandescent  lamps  must  necessarily  take  place  in  the  direction 
of  better  candle-power  maintenance ;  hence  this  phase  of  the 
problem  has  been  given  particular  study  during  the  past  two 
years. 

A  long  series  of  laboratory  experiments  led  to  the  introduction 
of  a  chemical  in  the  bulb  which,  under  proper  conditions,  will 
combine  with  the  black  deposit  in  such  a  manner  as  to  render  it 
light  in  color  and  thus  reduce  the  bulb  absorption  to  a  marked 
degree.  For  example,  the  lamp  whose  performance  is  shown  by 
curve  A  in  Fig.  2  has  a  life  sufficient  so  that  it  might  well  be 
operated  at  an  initial  efficiency  of  1  watt  per  candle  instead  of 
1.1  watt  per  candle,  its  actual  rating.  However,  if  burned  at  the 
former  efficiency,  its  candle-power  life  performance  would 
approximate  curve  B,  and  at  the  end  of  the  first  few  hundred 
hours  of  burning,  the  lamp  bulb  would  be  darkened  to  such  an 
extent  that  a  discriminating  user  would  become  dissatisfied  and 
would  complain  of  the  short  useful  life  of  the  lamp.  Others, 
not  so  careful,  would  perhaps  keep  the  lamp  in  service,  but  at 
the  expense  of  a  serious  loss  in  economy  due  to  the  constantly 
increasing  cost  of  energy  per  candle-power  hour.  By  the  use  of 
the  chemical  referred  to  above,  the  candle-power  life  perform- 
ance of  the  lamp  can  be  improved  in  accordance  with  curve  C 
and  its  useful  and  total  life  are  made  nearly  identical.  However, 
if  the  lamp  supplied  with  chemical  were  burned  under  voltage 
so  as  to  operate  at  an  efficiency  as  low  as  1.1  watts  per  candle,  its 
performance  would  not  be  that  shown  by  the  dotted  curve  D,  for 
the  .chemical  will  not  operate  properly  except  at  high  tempera- 
tures;  at  an  efficiency  of  1.1  watts  per  candle,  the  performance 
of  the  lamp  with  chemical  would  be  but  little,  if  at  all,  superior 
to  that  of  a  lamp  not  so  equipped. 

At  the  present  time  all  lamps  above  the  40-watt  size  are  sup- 


536  TRANSACTIONS  I.  £.   S. PART  II 

plied  with  chemical.  There  is  no  particular  advantage  in  intro- 
ducing the  chemical  into  smaller  lamps  inasmuch  as  the  useful 
life  is  still  limited  by  failure  of  the  filament  rather  than  by  bulb 
blackening. 

DECREASED  BULB  SIZE. 

Aside  from  bettering  the  life  performance  of  the  tungsten- 
filament  lamp,  much  time  and  effort  have  been  directed  toward 
decreasing  the  cost  of  the  product  and  broadening  its  application 
The  cost  of  manufacturing  and  handling  the  lamps  varies  almost 
directly  with  the  bulb  size,  and  the  cost  of  reflectors  and  similar 
accessories  increases  in  an  even  greater  ratio ;  hence,  from  the 
standpoint  of  economy,  a  substantial  decrease  in  bulb  size  is 
equivalent  to  a  considerable  increase  in  life  performance  or  effi- 
ciency. Fig.  3  shows  the  relative  sizes  of  the  old  60-watt  lamp 
with  the  skirted  base  and  the  lamp  as  at  present  marketed  in 
the  S-21  bulb.  Owing  to  the  use  of  chemical  in  the  new  60-watt 
lamp,  it  will  give  as  long  a  useful  life  as  the  older  type  at  a 
decrease  of  25  per  cent,  in  renewal  cost.  The  new  lamp  can  also 
be  used  in  many  locations  where  the  size  and  appearance  of  the 
older  type  rendered  it  inapplicable.  The  demand  for  a  60-watt 
tungsten-filament  lamp  in  a  smaller  bulb  was  probably  more  pro- 
nounced than  in  the  case  of  any  other  size,  inasmuch  as  this  is 
the  highest  wattage  which  would  be  used  to  replace  the  ordinary 
carbon  lamp  unit  for  unit.  At  the  same  time,  however,  there  is 
a  well-marked  tendency  toward  a  decrease  in  size  for  all  lamp 
bulbs. 

STANDARDIZATION. 

Of  particular  interest  to  the  illuminating  engineer  are  the  efforts 
which  have  been  put  forth  recently  toward  the  standardization 
of  lamp  dimensions.  Fig.  4  is  a  factory  specification  sheet  for 
the  new  60-watt  lamp,  for  which  forty-seven  distinct  items  have 
been  standardized.  The  most  important  standard  dimension  from 
the  viewpoint  of  the  illuminating  engineer  is  the  distance  between 
the  center  of  the  light  source  and  the  base  contact  of  the  lamp, 
since  it  is  this  dimension  which  has  the  greatest  effect  upon  the 
light  distribution  with  various  reflectors.  The  average  deviation 
from  the  standard  in  this  dimension  is  now  less  than  one-fourth 
of  the  average  deviation  found  a  year  ago. 


Fiar    6. 


[111! Will 


«T_ 


Fig.  7- 


Fig.  9. 


p      .  -  PPI 

^■t^SS::-.  J]"? 

"HffiSil            i&k 

HARRISON  AND  EDWARDS  :    INCANDESCENT  EAMPS  537 

HELICAL  FILAMENTS. 

Perhaps  the  most  far-reaching  of  the  improvements  which  have 
taken  place  during  the  past  year  is  the  general  introduction  of 
the  coil  filament  type  of  tungsten  lamp.  It  is  well  known  that 
after  the  voltage,  wattage  and  efficiency  of  an  incandescent  lamp 
have  been  determined,  all  of  the  filament  dimensions  are  fixed. 
For  example,  the  old  pressed-filament  i  io-volt,  40-watt  lamp  when 
designed  to  operate  at  1.25  watts  per  candle  had  a  filament  diam- 
eter of  1.605  mn"s  and  a  length  of  21.25  inches  (53.975  cm.). 
The  lamp  bulb  and  filament  supports  were  necessarily  of  such 
size  and  shape  as  to  take  care  of  this  length  of  filament.  It  was 
not  possible  to  obtain  a  concentrated  filament  lamp,  neither  could 
one  having  a  single  line  of  light  be  manufactured  unless  the  latter 
were  placed  in  a  bulb  22  inches  (55.88  cm.)  long.  Recently  it 
has  been  found  that  drawn  wire  filaments  of  all  sizes  can  be 
coiled  into  the  shape  of  a  helical  spring ;  this  greatly  reduces  their 
overall  length  and  makes  practicable  a  lamp  of  almost  any  form 
desired.  This  process  was  first  developed  in  connection  with  the 
low-voltage,  high-current  auto  headlight  lamps,  and  later  was 
found  practical  for  even  the  smallest  filament.  The  diameter  of 
the  helical  coil  is  ordinarily  not  more  than  seven  times  the  diam- 
eter of  the  filament  itself  and,  therefore,  the  difference  in  poten- 
tial between  successive  turns  is  very  small,  usually  about  one- 
tenth  of  a  volt.  At  this  low  voltage  there  is  practically  no  tend- 
ency for  the  current  to  short-circuit  its  regular  path,  even  when 
the  coils  appear  to  touch  each  other. 

In  addition  to  the  above,  it  has  been  found  that  coiled  filament 
lamps  are  much  stronger  than  those  of  the  standard  type,  quite 
as  strong,  in  fact,  as  the  old  carbon  lamps.  This  is  due  in  part 
to  the  greater  ability  of  the  coiled  filament  to  absorb  shock  and 
partly  because  tungsten  wire  seems  to  be  increased  in  strength 
by  stressing  beyond  its  elastic  limit. 

Fig.  5  shows  three  new  types  of  lamps  standardized  since  the 
introduction  of  the  coiled  filament.  Fig.  5-A  is  a  no-volt, 
15-watt  lamp  having  a  1  9/16-inch  (39.68  cm.)  bulb  with  cande- 
labra base,  and  illustrates  the  possibilities  in  small  high-efficiency 
light  sources. 

Fig.  5-B  illustrates  a  lamp  with  a  bulb  1  inch  (2.54  cm.)  in 
diameter  and   12  inches   (30.48  cm.)   long  that  is  made  in  the 


538 


TRANSACTIONS  I.  E.  S. — PART  II 


25  and  40-watt  sizes  and  is  intended  primarily  for  showcase 
lighting,  but  which  will  no  doubt  find  a  number  of  other  applica- 
tions. For  example,  Fig.  6  shows  this  lamp  in  a  frosted  bulb,  so 
placed  over  a  dresser  that  it  serves  as  a  satisfactory  substitute 
for  the  two  wall  brackets  often  located  at  either  side  of  the 
mirror. 

Fig.  7  shows  a  fixture  which  has  been  manufactured  for  use 
with  this  lamp  in  showcase  lighting.  Its  distinctive  features  are 
that  the  socket  is  supported  in  such  a  manner  that  the  lamp  can 
be  swung  into  the  position  shown,  in  order  to  facilitate  renewal 


175'     165°     155' 


25°         35°  45° 

REDUCTION  FACTOR 
100  WATT  REGULAR  Rt>.  BULB 
IOOWATT  FOCUS  FILAMENT  NO.  1 
100WATT  FOCUS  FILAMENT  NO.  2 

Fig.  8. 


81 
98 
89 


or  the  cleaning  of  the  reflectors;  that  the  current  is  carried 
through  successive  reflectors  by  dowel  pins  fitting  into  receptacles 
at  the  ends  of  each  unit  and  by  wires  placed  in  metal  moulding 
at  the  back ;  and  that  means  are  provided  for  securing  the  reflec- 
tor to  standard  showcases  of  all  types.  This  unit  is  assembled 
complete  before  leaving  the  factory  and,  therefore,  the  installa- 
tion can  be  made  at  a  minimum  labor  cost.    "Blanks"  or  "spacers" 


HARRISON  AND  EDWARDS  :    INCANDESCENT  LAMPS  539 

are  provided  where  the  high  intensity  which  would  be  obtained 
from  a  continuous  line  of  lamps  is  not  desired.  End  supports 
are  also  manufactured  to  carry  the  wires  up  from  the  floor  of 
the  showcase  and  to  secure  them  to  the  first  reflector. 

Fig.  5-C  shows  a  ioo-watt  concentrated  filament  lamp  of  the 
focus  type  which  should  fulfill  a  variety  of  needs.  The  coiled  fila- 
ment lamps  even  in  this  form  will  give  fully  as  good  a  life  per- 
formance as  lamps  of  the  regular  type  when  operated  under  simi- 
lar conditions.  In  most  cases,  however,  a  high  intensity  from  a 
very  small  source  is  desirable  and,  for  this  reason,  they  are 
operated  at  what  would  be  considered  over-voltage  for  the  stand- 
ard lamps  and  are  also  frequently  placed  in  smaller  bulbs  in  order 
to  permit  their  use  with  reflectors  and  lenses  of  moderate  dimen- 
sions. The  focal  length  of  a  lens  or  reflector  cannot  be  less  than 
one-half  the  diameter  of  the  bulb,  if  parallel  rays  of  light  are  to 
be  secured.  In  consequence  of  these  facts,  the  life  ratings  for 
focus  type  lamps  are  considerably  lower  than  for  the  correspond- 
ing standard  no-volt  units.  In  rating  coiled  filament  lamps  of 
all  types,  efficiency  and  candle-power  values  must  of  necessity  be 
based  on  total  light  output  rather  than  upon  the  intensity  in  any 
one  direction,  for  their  reduction  factor  is  much  higher  than  for 
ordinary  lamps  and  is  also  far  from  a  constant  quantity.  Fig.  8 
shows  the  distribution  curves  of  three  ioo-watt  lamps,  which 
illustrate  the  latter  point. 

The  ioo-watt  focus  type  lamp  will  successfully  operate  small 
stereopticons  for  lecture  room  or  residence  use  and  is  especially 
convenient  for  demonstration  work  in  connection  with  college 
courses.  It  is  perfectly  steady  and  requires  no  rheostat  and  no 
attention  other  than  switching  on  and  off.  With  these  lamps  it  is 
possible  to  obtain  a  brilliant  illumination  on  a  small  screen  for  a 
comparatively  low  energy  consumption. 

Fig.  9  illustrates  a  problem  in  sign  lighting  which  was  readily 
solved  by  means  of  concentrated  filament  lamps.  The  use  of  250- 
watt  regular  lamps  in  angle  reflectors  located  4  feet  out  from  the 
upper  edge  of  the  tank  was  first  considered,  but  it  was  found  that 
this  system  would  not  give  satisfaction  either  from  the  stand- 
point of  uniform  illumination  by  night  or  of  appearance  by  day. 
Two  ioo-watt,  no-volt  lamps  in  12-inch  (30.48  cm.)  silvered 
parobolic  reflectors  placed  in  a  weather-proof  housing  at  a  dis- 


540  TRANSACTIONS  I.  E.   S. — PART  II 

tance  of  approximately  50  feet  (15.24  m.)  supplied  the  illumina- 
tion for  the  portion  of  the  tower  illustrated  in  the  photograph. 
Two  more  units  are  used  to  illuminate  a  similar  sign  on  the  oppo- 
site side  of  the  tank. 

Fig.  10  shows  the  fagade  of  the  Engineering  Department  Build- 
ing, Cleveland,  Ohio,  as  it  will  appear  from  Euclid  Avenue  when 
illuminated  by  two  banks  of  projectors,  each  of  which  will  con- 
tain nine  250-watt  concentrated  filament  lamps  of  the  focus  type. 
Reflected  light  from  this  building  is  utilized  for  the  illumination 
of  the  grounds  immediately  adjoining. 

DISCUSSION. 

Mr.  Norman  Macbeth  :  There  is  one  point  I  would  like  to 
bring  in  here  in  connection  with  the  discussion  on  standardiza- 
tion which  arose  earlier  in  this  meeting.  It  is  agreed  that  the 
information  bearing  on  the  relative  position  of  the  filament  in 
an  incandescent  lamp  and  of  the  lamp  in  the  reflector  is  some- 
thing that  engineers  require  in  all  reports  of  photometric  inves- 
tigations. I  had  occasion  some  time  ago  to  go  over  the  standard 
specifications  of  four  laboratories,  and  in  checking  them  up  I 
found  that  their  "a,  b,  c,  d  and  e"  which  referred  to  these 
relations  showed  a  serious  lack  of  agreement  in  some,  one  or 
all  of  the  various  mentioned  designations. 

It  ought  to  be  the  duty  of  the  committee  of  this  society  to 
see  that  these  a,  b,  c,  d  and  e  designations  have  the  same  mean- 
ing with  all  of  these  laboratories,  and  it  would  also  be  desirable 
to  provide  for  other  glassware  and  fixture  conditions  than  those 
now  covered  by  these  letters  which  refer  only  to  simple  reflec- 
tor designations. 

I  would  like  to  raise  the  question  on  this  80  per  cent,  so- 
called  smashing  point  which  I  remember  Dr.  Carl  Hering 
stated  at  one  of  the  earlier  meetings  of  the  Philadelphia 
Section  he  had  determined  many  years  ago  by  himself, 
and  that  same  was  now  a  matter  of  record  in  the  proceedings 
of  the  American  Institute  of  Electrical  Engineers  This  was 
the  point  at  which  when  considering  an  average  cost  for  the 
lamps  and  for  the  energy  used,  a  lamp  having  deteriorated  to. 
80  per  cent,  of  the  initial  candle-power  was  no  longer  econom- 
ical ;  that  beyond  this  point  the  total  costs   in  relation  to  the 


INCANDESCENT    LAMPS  541 

light  produced  resulted  in  a  new  lamp  being  less  expensive. 
From  statements  that  have  been  made  applying  this  80  per  cent. 
to  the  tungsten  lamp  it  would  appear  that  we  have  lost  sight  of 
the  calculation  of  which  this  80  per  cent,  was  the  result.  It 
can  hardly  be  possible  that  with  the  more  expensive  tungsten 
lamps  and  the  very  much  lower  energy  cost  per  light  unit  that 
this  figure  would  be  true  for  these  lamps. 

I  would  like  to  ask  the  authors  the  probable  hour's  life  rep- 
resented by  the  curves  in  Fig.  2.  While  this  information  is  not 
really  necessary  to  bring  out  the  point  for  which  these  curves 
were  used  information  as  to  the  total  hours  represented  by  the 
two  curves  would  considerably  widen  the  use  of  this  one  illus- 
tration. 

Mr.  W.  F.  Little  :  The  standardization  of  filament  dimen- 
sions and  of  the  location  of  light  center  in  incandescent  lamps 
is  perhaps  of  more  importance  to  the  lamp  purchaser  than  to 
the  manufacturer,  as  the  variation  from  standard  conditions 
might  be  such  as  to  change  the  characteristic  of  an  intensive 
reflector  to  that  of  an  extensive  reflector.  As  representing  the 
purchaser,  the  Electrical  Testing  Laboratories  has  for  some 
time  past  urged  close  adherence  to  the  dimensions  as  laid  down 
by  the  manufacturer,  and  has  included  dimensional  requirements 
in  its  inspection  criteria. 

Mr.  G.  H.  Stickney  :  There  should  be  no  loss  in  efficiency 
from  light  falling  upon  filament  surfaces.  Any  heat  trans- 
mitted from  one  part  of  the  filament  to  another  in  this  manner 
would  simply  go  into  raising  the  temperature  of  the  filament 
and  therefore  be  returned  as  light.  (On  account  of  the  thin- 
ness of  filaments  and  the  heat  conduction  of  the  material,  it 
would  hardly  be  possible  for  one  side  of  the  wire  to  attain  a 
perceptibly  higher  temperature  than   the  other.) 

We  cannot  over-estimate  the  importance  to  the  art  which 
may  be  derived  from  our  ability  to  concentrate  the  filament,  as 
referred  to  in  the  paper.  As  referred  to  in  previous  discussion, 
this  has  introduced  a  considerable  number  of  new  applications 
for  incandescent  lamps  and  created  some  entirely  new  fields  of 
lighting.  The  importance  of  this  has  been  appreciated  for  some 
time  and  we  have  made  a  very  considerable  study  of  the  uses 


542  TRANSACTIONS  I.  E.   S. — PART  II 

of  point  sources  of  light.  In  the  long  run,  however,  I  am  not 
sure  that  the  point  source  may  not  prove  of  even  greater  im- 
portance in  handling  problems  of  interior  lighting,  since  it  offers 
an  opportunity  for  improved  control  of  the  light  in  obtaining 
new  and  desirable  effects. 

Mr.  L,.  C.  Porter:  In  speaking  of  the  applications  of  the 
new  focus  type  lamp,  there  are  three  fields  which  seem  to  open 
up  considerable  use  for  this  lamp.  One  is  in  theater  lighting. 
Some  of  the  larger  sized  lamps  have  been  used  for  flood  lights 
in  theaters.  By  the  use  of  such  lamps  and  large  parabolic 
reflectors  we  can  get  effects  which  are  hard  to  obtain  with  the 
arc.  The  incandescent  lamp  can  be  very  easily  and  evenly 
dimmed  or  brought  up  to  full  candle-power.  Focus  type  lamps 
have  been  used  with  parabolic  reflectors  to  throw  a  sheet  of  light 
over  each  individual  drop  curtain,  in  that  way  obtaining  much 
more  satisfactory  lighting  effects  than  previously  were  obtained 
by  the  use  of  the  arc.  The  light  is  steady  and  considerable 
power  economy  is  obtained  by  its  use. 

This  lamp  is  also  being  used  in  moving  pictures  for  home 
use,  to  do  away  with  the  fire  risk  and  auxiliary  apparatus  neces- 
sary with  the  arc. 

Another  field  in  which  a  little  experimenting  has  been  done 
is  in  signal  work.  Experiments  have  been  conducted  where 
signals  have  been  transmitted  over  a  distance  of  20  miles 
with  a  concentrated  filament  tungsten  lamp  and  parabolic  reflec- 
tor. These  signals  were  read  without  the  use  of  glasses,  both 
clear  and  colored  screens  being  used  over  the  reflectors. 

Still  another  field  which  is  opening  up  is  in  headlights,  espe- 
cially for  locomotive  headlights.  There  is  at  the  present  time 
considerable  agitation  throughout  the  country  on  the  question 
of  locomotive  headlights.  Several  states  have  passed  legisla- 
tion requiring  locomotives  to  carry  more  powerful  headlights ; 
in  others  this  subject  is  pending.  Some  of  the  roads,  especially 
those  with  double  tracks  equipped  with  block  systems,  object 
to  using  the  very  powerful  arc  headlights.  Between  the  arc 
and  the  oil  lamp  comes  the  incandescent.  It  seems  very  highly 
probable  that  a  6-volt  focus  type  of  incandescent  lamp  (oper- 
ated  either  by  storage  batteries   or  by   a   small   6-volt  turbo- 


INCANDESCENT   LAMPS  543 

generator    outfit)    will    have   large    application    for    locomotive 
headlights  in  the  near  future. 

Dr.  M.  G.  Lloyd:  The  efficiency  of  an  incandescent  fila- 
ment depends  upon  the  temperature  of  the  surface  which  is 
emitting  radiation.  The  question  arises  as  to  whether  with  a 
helical  filament  the  surface  temperature  can  be  as  high  as  with 
a  straight  filament  having  the  same  life;  or  whether  the  life 
can  be  as  long  for  the  same  surface  temperature.  In  a  straight 
filament  the  highest  temperature  is  in  the  interior  of  the  wire 
and  it  is  only  the  temperature  of  the  surface  which  affects  either 
radiation  or  evaporation.  In  a  helical  filament  the  highest  tem- 
perature will  probably  be  found  on  the  surface  of  the  wire  in 
the  interior  of  the  coil.  This  surface  is  exposed  sufficiently  to 
permit  evaporation,  but  not  sufficiently  to  give  out  radiation. 
The  temperature  of  the  exterior  surface  will  as  before  deter- 
mine the  efficiency;  whereas,  the  temperature  of  the  interior 
surface  will  probably  determine  the  life.  For  the  same  tem- 
perature of  the  external  surface  it  would,  consequently,  seem 
possible  that  the  evaporation  would  be  greater,  and  hence  the 
life  of  the  filament  shortened  as  compared  with  a  straight  fila- 
ment.* Practical  experience  seems  to  indicate  that  this  effect 
is  not  appreciable. 

Mr.  H.  Calvert  :  Referring  to  Fig.  2,  I  note  that  the  figures 
representing  the  hours  life  have  been  omitted.  I  think  it  would 
be  much  more  interesting  and  instructive  if  these  figures  could 
be  inserted.  The  question  of  the  blackening  of  the  lamp  bulbs 
has  recently  taken  on  a  new  and  interesting  phase.  A  number 
of  the  central  stations  have  recently  adopted  the  policy  of  giving 
free  renewals  on  certain  sizes  of  tungsten  lamps.  Formerly, 
when  the  consumer  had  to  purchase  each  lamp  he  would  keep 
it  in  use  until  it  got  so  black  that  he  could  get  but  little  light. 
Now,  under  the  new  ruling,  he  is  entitled  to  free  renewals,  and 
the  question  arises,  at  what  point  in  the  life  of  the  lamp  is  the 
company  justified  in  giving  a  new  lamp?  It  is  generally  not  so 
much  the  decrease  in  the  actual  candle-power  of  the  lamp  which 
influences  him  in  bringing  the  old  lamp  to  the  central  station,  as 

*  Dr.  Irving  t,angmuir  has  since  stated  that  in  the  half-watt  lamp  using  the  helical 
form  of  filament  the  interior  surface  is  at  a  temperature  about  12  degrees  higher  than  the 
exterior  surface. 


544  TRANSACTIONS  I.  $.   S. PART  II 

it  is  the  blackened  appearance  of  the  bulb.  I  would  therefore 
like  to  ask  the  authors  what,  in  their  opinion,  is  the  limit  of  this 
term  "useful  life"  which  they  use,  and  at  what  point,  in  their 
opinion,  are  the  various  companies  justified  in  giving  free  re- 
newals ? 

Dr.  R.  E.  Myers  :  This  is  a  very  interesting  paper  which  the 
authors  have  given  us  and  there  is  very  little  to  add.  The  prin- 
cipal features  upon  which  the  recent  improvements  depend  have 
been  given  in  a  very  able  way. 

Greater  emphasis  might  be  laid  on  the  fact  that  the  latest 
developments  in  the  standard  lines  of  tungsten  lamps,  which  prob- 
ably represent  over  90  per  cent,  of  this  product,  are  due  to 
improvements  in  the  filament  and  to  the  use  of  chemicals. 

The  improvements  in  filament  are  of  two  kinds.  First,  we  are 
now  using  a  drawn  wire  which  has  a  much  greater  tensile 
strength  than  the  pressed  filament  of  the  earlier  tungsten  lamp. 
This  gives  low  manufacturing  shrinkage  and  reduced  cost. 
Second,  the  present  filament  is  much  stronger  throughout  its 
burning  than  the  older  filament.  This  is  due  to  certain  changes 
of  which  I  am  not  at  liberty  to  speak  owing  to  trade  reasons. 
These  changes,  however,  are  of  the  utmost  importance  in  the  life 
of  the  lamps  of  lower  wattage  of  the  standard  line. 

Large  improvements  in  efficiency  have  certainly  been  made 
by  the  use  of  chemicals  in  the  lamp.  However,  I  believe  that 
still  greater  ones  are  yet  to  come.  This  field  of  research  will 
doubtless  prove  a  very  fertile  one  and  I  think  that  in  a  great 
many  cases  this  type  of  lamp  in  actual  use  will  prove  equal  to  or 
even  superior  to  the  new  gas  filled  lamps. 

The  authors  may  have  to  change  their  opinion  regarding  the 
use  of  chemicals  in  the  smaller  lamps.  Recent  experiments  tend 
to  prove  that  certain  chemicals  can  be  used  to  advantage  in  these 
also. 

Dr.  H.  E.  Ives  :  Another  application  of  these  helical  fila- 
ments— not  a  commercial  application — is  in  photometric  research. 
In  this  a  valuable  consideration  is  the  ability  to  obtain  a  wide 
range  of  illumination  without  resorting  to  sector  disks  or  absorb- 
ing screens.  With  the  ordinary  carbon  lamp  it  is  impossible  to 
get  nearer  the  photometric  screen  than  about  10  inches,  because 


INCANDESCENT   LAMPS  545 

the  inverse  square  law  can  no  longer  be  applied.  If,  however 
the  light  is  compressed  into  these  little  coils  it  is  possible  to 
push  the  light  up  until  the  bulb  actually  strikes  the  receiving 
surface  without  involving  errors  from  inverse  square  law  cal- 
culations which  means  that  one  can  increase  the  range  of  il- 
lumination at  least  ioo  times  over  what  was  available  before. 
In  this  connection  I  have  had,  through  the  kindness  of  Mr. 
Stickney,  several  of  these  lamps  to  try  out  and  I  have  found 
them  extremely  convenient  in  so  far  as  the  ability  to  obtain  great 
variation  of  illumination  is  concerned. 

One  question,  however,  I  have  not  been  able  to  investigate.  I 
would  like  to  ask  the  authors  of  this  paper  as  to  the  performance 
of  these  lamps.  I  suspect  that  the  high  voltage  lamp  would  be 
apt  to  be  unsteady  and  unreliable  because  of  the  number  of  fila- 
ment supports.  The  low  voltage,  however,  might  be  very  reliable 
indeed.  I  should  like  to  ask  the  authors  if  they  have  any  data 
that  might  enable  us  to  make  comparisons  as  to  steadiness  with 
the  usual  photometric  standards. 

Mr.  J.  R.  Cravath  :  I  would  like  to  ask  what  the  effect  of 
coiling  the  filament  in  helical  form  has  upon  the  efficiency.  Is 
there  any  loss  of  efficiency  due  to  the  shading  effect  of  one  con- 
volution on  the  next? 

Mr.  V.  R.  Lansingh  :  The  use  of  the  helical  coil  filament 
giving  a  very  small  body  of  light  will  be  particularly  useful  in 
the  design  of  reflectors  using  specular  reflection,  such  as  pris- 
matic, mirrored,  etc.  One  of  the  greatest  difficulties  in  a  design 
of  that  character  is  the  size  of  the  source  of  light.  The  ordi- 
nary tungsten  lamp  filament  may  be  considered  as  a  trans- 
parent cylinder,  and  this  introduces  a  number  of  problems,  one 
of  which  is  the  shape  of  prisms,  etc.,  in  the  case  of  prismatic 
glassware.  The  ordinary  prism,  as  you  probably  all  know,  is 
90  degrees.  It  is  possible,  however,  to  design  prisms  of  differ- 
ent shapes,  so  that  a  higher  efficiency  is  obtained  with  the  pres- 
ent type  lamp  filament  than  given  by  the  90  degree  prism.  For 
example,  one  manufacturer  uses  a  parabolic  prism  on  the  out- 
side of  the  reflector. 

With  the  introduction,  however,  of  the  new  filament,  it  will 
be  possible  to  obtain  far  higher  efficiency  than  before,  and  this 


54^  TRANSACTIONS  I.  £.   S. — PART  II 

will  be  particularly  valuable  where  broad  distributions  of  light 
are  wanted,  such  as  in  street  lighting.  We  may  therefore  look 
probably  for  a  large  advance  in  the  design  of  reflectors  from 
now  on  with  the  new  filament  lamp. 

Mr.  L.  J.  Lkwinson  :  Referring  once  again  to  the  light 
colored  discoloration  due  to  the  chemical,  I  would  like  to  ask 
the  authors  whether  the  manufacturers  consider  it  necessary 
to  ship  from  the  factory  new  lamps,  the  bulbs  of  which  already 
manifest  this  discoloration  to  some  extent. 

Mr.  H.  S.  Dunning  :  Just  a  point  as  to  the  return  of  lamps 
to  central  stations :  It  is  probable  that  for  some  time  lamps  of 
the  150-watt  or  larger  sizes  may  be  returned  because  they  show 
a  white  deposit  on  the  bulb.  It  has  been  our  experience  backed 
by  accurate  measurements  that  in  most  cases  the  appearance  of 
this  white  deposit  does  not  necessarily  mean  a  large  decrease 
in  candle-power,  and  quite  often  it  will  be  found  that  the 
candle-power  has  not  decreased  at  all.  Some  of  these  lamps 
have  a  most  interesting  performance  on  life  test.  We  have 
found  cases  in  which  lamps  after  operating  for  an  extended 
period  at  approximately  100  per  cent,  of  their  initial  candle- 
power  seem  to  lose  candle-power  for  a  time  and  later  come  back 
to  approximately  normal  rating.  These  changes  have  been 
investigated  very  closely  and  it  has  been  found  that  they  are 
not  due  to  either  photometric  or  other  test  errors.  I  think, 
therefore,  that  a  word  of  caution  is  in  order  against  the  dis- 
carding of  such  lamps,  simply  because  they  show  an  unusual 
discoloration  of  the  bulb. 

Dr.  M.  G.  Lloyd:  The  statement  of  Mr.  Harrison  that  the 
smashing  point  does  not  depend  upon  the  relative  cost  of  energy 
and  renewals  involves  an  assumption  that  should  be  made  clear. 
It  does  not  apply  to  lamps  operated  at  rated  voltage,  since  it  is 
clear  that  if  the  cost  of  renewal  was  merely  nominal,  the  smash- 
ing point  would  be  reached  as  soon  as  the  candle-power  has 
fallen  off  appreciably,  say  to  95  per  cent.  Mr.  Harrison's 
assumption  is  that  the  lamp  shall  be  operated  at  the  most  eco- 
nomical efficiency  and  the  lower  the  cost  of  renewals  the  lower 
should  be  the  specific  consumption  and  the  higher  the  voltage 
at  which  a  given  lamp  should  be  operated.     Since  this  is  a  con- 


INCANDESCENT   LAMPS  547 

dition  that  cannot  be  conveniently  carried  out  in  practise,  the 
statement  that  the  smashing  point  is  independent  of  the  cost  of 
renewals  and  the  cost  of  energy  is  also  one  that  does  not  apply 
to  practical  conditions. 

Mr.  Ward  Harrison  (In  reply)  :  Mr.  Lewison  inquires  in 
what  way  the  characteristics  of  the  coiled  filament  lamps  differ 
from  those  having  the  standard  form  of  filament.  There  is  a 
slight  variation  in  the  performance  curves,  but  the  principal  dif- 
ference is  in  the  manufacturing  data ;  i.  e.,  the  length  of  filament 
required,  and  the  diameter  for  a  given  candle-power,  voltage  and 
efficiency. 

The  abscissae  of  the  curves  in  Fig.  2  were  not  indi- 
cated as  actual  hours  of  life  because  the  relative  performance 
with  a  given  change  in  voltage  is  independent  of  the  actual 
hours  represented  by  the  curve.  A  typical  lamp  giving  2,000 
hours  life  to  burn  out  without  "getter"  would  have  a  life  of 
about  1,000  hours  if  the  voltage  were  raised  10  per  cent.  Under 
those  conditions  the  introduction  of  the  "getter"  would  result 
in  better  candle-power  maintenance  but  the  life  to  burn  out 
would  still  be  about  1,000  hours.  Actual  quantitative  data  on 
actual  lamp  performance  were  given  by  Mr.  Lewinson. 

It  has  long  been  considered  that  the  smashing  point  of  incan- 
descent lamps  for  most  economical  operation  depends  upon  the 
rate  for  energy  and  the  cost  of  lamps,  as  well  on  the  form 
of  the  candle-power  depreciation  curve.  More  recent  investi- 
gations show  that  this  smashing  point,  a  certain  percentage  of 
initial  candle-power,  is  determined  solely  by  the  performance  of 
the  lamp.  The  most  economical  life  of  the  lamp  in  hours,  on 
the  other  hand,  is  determined  by  the  cost  of  lamp  renewals  and 
the  energy  rate,  and  this  life  can  be  secured  in  practise  by 
selecting  a  lamp  of  the  correct  efficiency.  If  the  proper  choice 
of  initial  efficiency  is  made,  the  lamp  will  be  found  to  have 
dropped  in  candle-power  to  the  smashing  point  at  the  end  of  the 
period  representing  its  economical  life  as  determined  above.  A 
full  discussion  of  this  subject  is  given  in  a  technical  bulletin 
issued  by  the  engineering  department,  National  Lamp  Works 
of  the  General  Electric  Company. 

The  answer  to  the  question  raised  by  Dr.  Ives  in  regard  to 


54§  TRANSACTIONS  I.  E.   S. — PART  II 

using  concentrated  filament  lamps  as  standards  is  suggested  by 
Fig.  8  of  the  paper.  Curves  i  and  2  are  both  on  concentrated 
filament  lamps  built  to  fulfill  the  same  specifications  and  yet  it 
is  seen  that  the  candle-power  curves  are  considerably  different. 
Inasmuch  as  a  very  slight  change  in  the  relative  position  of 
filaments  will  cause  a  marked  change  in  the  shading  effect  such 
lamps  can  scarcely  be  used  as  standards.  The  filament  may 
sag  slightly  after  the  lamp  has  been  burning  for  a  short  period 
and  thus  materially  change  the  distribution  curve,  although  the 
total  light  flux  and  efficiency  would  not  be  affected  to  any  extent. 
This  well  illustrates  the  necessity  of  arriving  at  something 
better  than  simply  a  watt  per  horizontal  candle-power  rating  for 
incandescent  units. 

Mr.  Macbeth's  remarks  were  most  timely  as  to  the  desirability 
of  having  some  uniform  system  of  designating  the  position  of 
the  lamp  filaments  relative  to  the  reflector  in  recording  data  on 
photometric  tests.  I  am  glad  to  say  in  this  connection  that  we 
have  taken  the  matter  up  with  several  other  laboratories  and 
have  agreed  upon  a  system  of  lettering  which  will  be  uniform 
and  which  no  doubt  will  eliminate  much  of  the  confusion  which 
has  existed  heretofore.  These  designations  are  of  course  purely 
arbitrary. 

Mr.  Cravath  and  Dr.  Lloyd  spoke  about  the  rise  in  tempera- 
ture of  the  inner  surface  of  the  filament  because  of  the  tend- 
ency of  one  portion  of  the  coil  to  shade  another.  This  effect 
so  far  as  has  been  determined  is  very  slight.  As  Mr.  Stickney 
has  said,  if  the  radiant  energy  does  not  escape  at  once,  it  is 
merely  reflected  from  the  hot  surface  of  the  filament  one  or 
more  times  before  passing  out  of  the  helix.  If  this  radiant, 
energy  would  be  absorbed  instead  of  being  reflected,  it  would 
of  course  raise  the  filament  temperature  somewhat,  but  the  net 
change  would  certainly  be  very  slight,  since  the  difference  in 
temperature  between  two  points  on  the  same  cross  section  of 
filament  is  necessarily  small. 

There  is  one  more  point,  and  that  is  in  regard  to  the  dis- 
coloration of  the  bulb  of  "getter"  lamps  without  a  material 
decrease  in  light  output,  which  was  commented  upon  by  one  of 
the   speakers.      This   phenomenon    is   entirely   reasonable   as    in 


INCANDESCENT    LAMPS  549 

those  cases  the  deposit  is  light  in  color,  similar  in  fact  to  an 
opal  dip  or  frosting.  It  may  of  course  have  a  considerable 
effect  on  the  intensity  of  the  lamp  in  any  one  direction,  but  the 
decrease  in  total  light  flux  should  not  exceed  a  few  per  cent. 
The  tendency  of  these  lamps  to  be  slightly  erratic  is,  of  course, 
not  surprising;  however,  we  believe  that  the  improvement  in 
this  respect  during  the  past  few  months  has  been  marked. 

Mr.  J.  W.  Howell:  Most  of  the  things  that  I  had  in  mind 
to  speak  about  have  already  been  mentioned  either  by  Mr.  Harri- 
son and  Mr.  Edwards  or  in  the  report  of  your  Committee  on 
Progress,  so  all  I  can  do  will  be  to  give  you  a  little  further  in- 
formation on  some  of  the  things  which  you  have  already  been 
considering.  In  the  time  in  which  we  live  the  most  rapid  advance 
in  the  art  of  light  producing  ever  known  is  being  made.  Even 
since  incandescent  lamps  have  been  made  one  of  their  character- 
istics has  been  that  the  lamps  with  thin  filaments  end  their  life 
by  breaking,  without  much  discoloration  of  the  bulb,  while 
lamps  with  thick  filaments  and  high  candle-power  blacken  the 
bulb  and  become  useless  before  the  bulb  breaks.  So  we  have 
never  been  able  to  obtain  the  good  results  which  we  should  get 
from  thick  filament  lamps.  Much  progress  has  been  made  in 
the  last  two  years  in  preventing  the  blackening  of  the  bulbs  and 
thus  utilizing  the  longer  life  of  the  thick  filament.  A  year  ago 
we  talked  and  discussed  the  action  of  what  we  call  vacuum 
getters  on  lamps.  That  same  line  of  work  has  been  conducted 
during  the  past  year  and  very  great  advances  have  been  made. 

There  are  two  reasons  why  a  thick  filament  lamp  gets  blacker 
than  a  thin  filament  lamp :  one  of  the  reasons  is  that  it  lives 
longer  and  consequently  has  more  time  to  get  black  in ;  the  sec- 
ond is  the  blackening  is  proportional  to  the  surface  of  the  fila- 
ment, and  inversely  to  the  size  of  the  bulb.  To  be  properly  de- 
signed the  surface  of  the  lamp  bulb  should  be  proportional  to 
the  surface  of  the  filament.  But  that  cannot  be.  A  io-watt  lamp 
has  a  diameter  of  2%  inches,  that  would  require  a  100-watt  lamp 
to  be  10  times  that  diameter  of  21  inches;  which  would  be  im- 
possible. 

During  the  year  we  have  made  very  great  advances  in  these 
large  lamps. 


550  TRANSACTIONS  I.  £.  S. — PART  II 

I  have  just  shown  a  report  of  some  life  tests  on  some  hundred- 
watt  hundred  volt  lamps.  These  lamps  are  tested  at  9/ioths  of 
a  watt  per  candle.  Their  normal  burning  efficiency  is  one  watt 
per  candle.  Of  these  tests  only  two  were  completed  at  the  time 
the  table  was  made  up,  which  is  quite  recently.  In  the  two 
of  the  tests  the  lamps  burned  a  thousand  hours  at  9/ioths  of  a 
watt  per  candle  and  at  the  end  of  that  time  were  up  to  their 
initial  performance.  At  the  end  of  3,100  hours  they  should  show 
99  per  cent,  of  their  full  candle-power. 

I  also  have  a  report  of  a  test  of  250-watt  100-volt  lamps.  The 
lamps  were  not  experimental  lamps ;  they  are  not  made  in  the 
laboratory ;  every  lamp  was  taken  from  the  regular  factory  stock, 
taken  without  any  selection  whatever.  The  lamps  were  burned 
at  24  of  a  watt  per  candle  on  life  tests.  After  500  hours  burn- 
ing they  were  pretty  close  to  90  per  cent,  candle-power  of  initial 
candle-power.  If  they  had  been  burned  at  their  normal  effi- 
ciency, which  is  one  watt  per  candle,  their  life  would  have  been 
9.6  as  long.  So  that  when  they  were  down  to  90  per  cent,  of 
their  candle-power,  burning  on  a  normal  voltage,  they  would 
have  burned  over  4,000  hours.  That  shows  very  great  improve- 
ment in  lamps  of  that  type. 

I  also  have  a  diagram  showing  a  test  on  a  400-watt  lamp  taken 
right  out  of  stock,  without  any  selection  whatever.  The  lamp 
was  started  on  test  in  September  of  last  year,  just  a  year  ago. 
When  the  diagram  was  made  out,  the  lamp  had  burned  6,000 
hours  and  showed  about  92.5  per  cent,  of  its  initial  candle-power. 
The  lamp  when  I  left  to  attend  this  convention  had  burned  7,000 
hours,  practically  unchanged  at  this  date. 

These  reports  indicate  the  great  improvement  which  has  been 
made  in  the  high  candle-power  lamps.  It  is  the  established  pro- 
cedure in  business  that  as  improvements  are  made  the  efficiency 
is  increased.  Otherwise  the  economy  of  the  lamp  would  not  be 
realized.  It  is  considered  at  the  present  that  a  thousand  hours 
of  useful  commercial  life  is  proper  and  best,  so  that  as  lamps 
come  to  give  long  lives  like  that  their  efficiency  is  increased  so 
that  the  laboratory  life  is  about  1,300  or  1,400  hours,  which  would 
give  in  commercial  practice  a  life  of  about  1,000  hours. 

There  has  been  under  way,  during  the  last  two  years  a  very 


INCANDESCENT   LAMPS  551 

remarkable  piece  of  work  in  the  research  laboratory  of  the  Gen- 
eral Electric  Company  at  Schenectady.  The  work  has  been  done 
by  Dr.  Langmuir  and  his  assistants,  one  of  whom  is  here  today, 
and  it  was  my  purpose  to  tell  you  a  good  deal  about  the  work  at 
this  meeting  but  unfortunately  Dr.  Langmuir's  description  has  had 
to  be  postponed  until  next  month.  A  half-watt  per  candle  lamp, 
— that  lamp  is  the  result  of  Dr.  Langmuir's  work.  It  is  a  lamp 
such  as  we  have  here,  an  incandescent  tungsten  filament  lamp, 
the  bulb  of  which  is  filled  with  nitrogen.  Now  the  nitrogen  has 
several  effects  on  the  lamp ;  some  are  good  and  some  are  bad. 
The  bad  one  is  that  it  cools  the  filament  and  so  reduces  the  candle- 
power.  When  a  gas  is  introduced  into  a  lamp  vacuum  it  cools 
the  filament  always.  In  a  thin  filament  lamp  the  cooling  effect 
is  greater  than  in  thick  filament  lamps.  So  the  thicker  the  fila- 
ment the  more  benefit  is  to  be  had  from  the  nitrogen  gas  in  the 
bulb.  In  the  present  state  of  our  knowledge  we  get  our  one  half- 
watt  per  candle  lamp  at  about  12  amperes;  while  at  20  amperes 
we  get  0.4-watt-per  candle.  Tests  of  such  20  ampere  lamps,  have 
been  made  at  0.4-watt  per  candle.  Some  of  these  lamps  have 
burned  2,000  hours.  The  large  lamp  here  before  you  is  a  20-am- 
pere  112-volt  lamp;  it  consumes  about  2,500  watts,  and  gives 
5,000  candle-power.  It  is  supposed  to  be  the  largest  incandescent 
lamp  which  has  ever  been  made.  Unfortunately  the  lamp  has  a 
little  crack  in  the  glass,  which  has  allowed  a  little  air  to  mix  with 
the  nitrogen ;  so  that  when  it  is  lighted  it  will  show  a  faint  cloud 
of  tungsten  oxide,  and  it  is  not  an  effective  lamp.  The  future 
of  that  lamp,  gentlemen,  you  may  all  theorize  on  as  much  as  I 
can.  You  know  what  the  introduction  into  the  art  of  lamps  of 
larger  sizes  than  have  been  made  before  means ;  and  the  policy 
has  been  established  to  add  to  our  present  line  of  lamps  up  to 
this  5,000  candle-power  size,  or  higher  if  necessary.  If  the 
commercial  people  want  a  10,000  candle-power  lamp  they  can 
have  it ;  there  is  no  limit  that  we  know  of  yet. 

That  lamp  which  you  see  burning  is  a  500  candle-power  lamp, 
at  112  volts.  Its  light  is  very  concentrated  and  very  intense.  The 
lamp  is  of  a  size  which  does  not  realize  a  half-watt  per  candle: 
in  fact  it  is  about  a  0.6  watt  per  candle  lamp. 

Here  is  a  larger  lamp  one  which  burns  at  one-half -watt  per 


552  TRANSACTIONS  I.  E.   S. — PART  II 

candle ;  it  is  a  good  lamp.  The  air  is  leaking  into  that  bulb ;  its 
action  will  not  continue  very  long,  but  while  it  is  continuing  it 
is  destroying  the  filament.  I  don't  know  what  will  happen  to 
the  lamp  with  the  air  in  it. 

There  has  been  another  matter  which  has  been  alluded  to  in 
the  report  of  the  Committee  on  Progress  and  also  in  the  paper, 
by  Messrs.  Harrison  and  Edwards,  which  I  consider  very  im- 
portant and  that  is  what  they  call  a  single  size  wire.  It  simply 
means  that  the  art  of  drawing  tungsten  wire  has  been  reduced  to 
such  a  fine  degree,  such  an  efficiency,  that  the  wire  can  be  drawn 
to  absolutely  the  size  desired.  Of  course  when  you  consider  the 
matter  if  the  dies  are  proper  and  of  the  right  size  the  wire 
drawn  should  be  the  right  size.  They  are  the  right  size. 
And  at  the  present  the  wire  in  the  filament  is  of  the 
right  length  and  diameter;  all  of  the  lamps  made  for  a  given 
voltage  are  that  voltage.  If  a  factory  is  producing  lamps  that 
are  of  a  given  voltage  and  the  photometer  disagrees  with  the 
marking,  the  photometer  is  wrong.  (Laughter.)  Gentlemen, 
that  is  a  fact.  It  is  true  that  the  grading  is  better  than  the  photo- 
meter. If  we  find  that  the  lamps  test  off  voltage,  in  nine  cases 
out  of  10  the  difficulty  is  with  the  photometer  and  not  with  the 
lamps.  As  you  know  some  of  our  customers  in  the  country  buy 
their  lamps  on  specifications  as  to  voltage  and  candle  power,  and 
the  lamps  submitted  during  the  last  six  months  or  more  to  those 
customers  are  lamps  which  have  never  been  photometered ;  they 
are  lamps  made  for  a  definite  voltage  and  candle  power,  and  the 
result  is  that  they  are  closer  to  rating  than  were  the  previous 
photometered  lamps. 

Mr.  L,.  J.  Lewinson  :  In  order  to  further  emphasize  some 
of  the  points  brought  up  in  the  paper,  a  table  and  a  diagram  are 
submitted  herewith.  The  table  comprises  a  summary  of  the 
various  watts  per  candle  ratings  of  tungsten  lamps  in  force  in 
this  country  during  the  past  two  years.  It  is  noted  that  with 
the  exception  of  the  very  smallest  sizes  great  strides  in  efficient 
improvement  have  been  made.  The  diagram  shows  the  very  re- 
markable improvement  life  and  efficiency  of  one  particular  size — 
the  250-watt  lamp.  The  heavy  horizontal  lines  represent  the 
average  life  value  during  the  periods  indicated.     It  is  seen  that 


INCANDESCENT    LAMPS 


553 


Efficiency  Adjustment— Tungsten  Lamps. 

Watts-per-candle  rating- 
Wattage  April         September         May 
ratings                                              191 1  1912  1912  1913 

15  watts 1. 31  1. 31  1.30  1.30 

20      "      1. 31  1.28  1.25  I.25 

25  "  I-31  1-23  I.I7  I-I7 

40  "  1.23  1. 18  1. 17  1. 17 

60  "  1. 18  1. 16  1. 16  1. 12 

100  "  1. 18  1. 13  1. 13  1.08 

150  "  1. 18  1. 12  1. 12  1.03 

250  " 1. 13  1. 10  1. 00  1. 00 

400  "  1. 13  1. 10  1. 00  1. 00 

500  "  1. 13  1. 10  1. 00  1. 00 

early  in  191 1  the  useful  life  of  these  lamps  was  about  700  hours 
at  1. 1 3  watt-per-candle.  In  October  of  the  same  year  a  bulb  black- 
ening preventive  was  introduced  which  had  the  effect  of  increas- 


— 1 

WRC 

W.RC.        . 

1.13 

W.P.C. 

<U0_ 

,          1.00 

"* 

^ 

\ 

JAN/12 


JULY'15 


Diagram  A.— Improvement  in  useful  life  of  250-watt  tungsten  lamps. 

ing  the  life  to  1,350  hours.  By  May,  1912,  the  manufacture  of 
this  type  of  lamp  had  been  so  improved  that  it  was  possible  to 
raise  the  efficiency,  lowering  the  watts  per  candle  to  1.10,  with- 
out deleterious  effect  upon  the  life.  In  July,  1912,  a  new  bulb 
blackening  preventive  was  adopted,  increasing  the  life  to  nearly 
2,100  hours.  In  September  of  the  same  year  a  new  form  of  con- 
struction was  adopted.    It  was  found  possible  to  still  further  in- 


554  TRANSACTIONS  I.  E.   S. — PART  II 

crease  the  efficiency,  to  correspond  to  i.o-watt  per  candle  with 
only  a  very  slight  loss  in  life.  In  February,  191 3,  the  form  of 
construction  was  again  changed,  with  a  resultant  life  of  2,500 
hours.  It  is  seen  then  that  during  the  brief  span  of  two  years 
the  watts  per  candle  rating  has  been  reduced  from  1.13  to  1.10 
with  an  increase  of  250  per  cent,  in  useful  life.  Assuming  all 
life  values  to  be  corrected  to  one  standard  watt  per  candle  value, 
the  inherent  improvement  in  the  life  of  this  particular  type  of 
lamps  has  been  nearly  700  hours.  All  life  values  quoted  above 
are  to  be  considered  as  life  to  80  per  cent,  or  prior  burn-out. 


Illuminating  Engineering  Society 

GENERAL  OFFICES :     29  West  39th  Street,  New  York  City. 

SECTIONS 


Chicago  Section 

CHAIRMAN M.  G.  Lloyd,  608  S.  Dearborn  Street,  Chicago,  111. 

SECRETARY J.  B.  Jackson,  28  N.  Market  Street,  Chicago,  111. 

MANAGERS 

Nelson  M.  Black 1213  Wells  Building,  Milwaukee,  Wis. 

J.  W.  Pfiefer 520  Hannah  Avenue,  Forest  Park,  111. 

C.  C.  Schiller 122  Michigan  Boulevard,  Chicago,  111. 

M.  J.  Sturm 116  S.  Michigan  Boulevard,  Chicago,  111. 

H.B.Wheeler 6204  Lakewood  Avenue,  Chicago,  111. 


New  England  Section 

CHAIRMAN  .    .  C.  A.  B.  Halvorson,  Jr.,  G.  E.  Co.,  Center  Street,  West  Lynn,  Mass. 
SECRETARY  .  Chas.  M.  Cole,  156  Pearl  Street,  Boston,  Mass. 
MANAGERS 

R.  B.  Hussey  ....  General  Electric  Company,  West  Lynn,  Mass. 

H.  C.  Jones 10  High  Street,  Boston,  Mass. 

J.M.Riley East  Boston  Gas  Company,  Chelsea,  Mass. 

R.  C.  Ware 42  West   Street,  Boston,  Mass. 

W.  E.  Wickenden,  Massachusetts  Inst,  of  Technology,  Boston,  Mass. 


New  York  Section 

CHAIRMAN  ....  William  Cullen  Morris,  124  East  15th  Street,  New  York,  N.  Y. 
SECRETARY    .   .   .  Clarence  L.  Law,  124  West  42nd  Street,  New  York,  N.  Y. 
MANAGERS 

H.B.Rogers General  Electric  Company,  Harrison,  N.  J. 

A.  S.  Ives 84  William  Street,  New  York,  N.  Y. 

H.  V.  Allen     13  Park  Row,   New  York,  N.  Y. 

W.  H.  Spencer 239  Tenth  Avenue,  New  York,  N.  Y. 

Oscar  H.  Fogg 124  East  15th  Street,  New  York,  N.  Y. 


Philadelphia  Section 

CHAIRMAN     .   .  George  A.  Hoadley,  Swarthmore  College,  Swarthmore,  Pa. 
SECRETARY   .   L.  B.  Eichengreen,  Counties  Gas  &  Elec.  Co.,  Ardmore,  Pa. 

MANAGERS 

H.Calvert 1000  Chestnut  Street,  Philadelphia,  Pa. 

F.  C.  Dickey 30  S.  16th  Street,  Philadelphia,  Pa. 

H.  H.  Ganser      .  212  DeKalb  Street,  Norristown,  Pa. 

H.  A.  Hornor  .    .  102  Hamilton  Court,  39th  and  Chestnut  Streets, 

Philadelphia,  Pa. 
Samuel  Snyder 1035  Market  Street,  Philadelphia,  Pa. 

Pittsburgh  Section 

CHAIRMAN C.   J.    Mundo,    1312  Oliver  Building,  Pittsburgh,  Pa. 

SECRETARY Alan  Bright,  827  Wabash  Building,   Pittsburgh,  Pa. 

MANAGERS 

H.  S.  Hower Carnegie  Technical  Schools,  Pittsburgh,  Pa. 

H.  H.  Magdsick Nela  Park,  Cleveland,  Ohio. 

E.  R.  Roberts S40  Middle  Street,  Avalon,  Pa. 

C.  E.  Stephens  .  Westinghouse  Elec.  &  Mfg.  Co.,  E.  Pittsburgh,  Pa. 
S.  B.  Stewart 435  Sixth  Avenue,  Pittsburgh.  Pa. 


Illuminating  Engineering  Society 


LIST  OF  LOCAL  REPRESENTATIVES 


State  and  City  Name  and  Address  of  Representative 

CALIFORNIA:    Los  Angeles R.  H.  Manahan, 

City  Electrician,  Los  Angeles,  Cal. 
San  Francisco    .   .   .   .  F.  Emerson  Hoar, 

833  Market  Street,  San  Francisco,  Cal. 
COLORADO:       Denver G.E.Williamson, 

Denver  Gas  &  Electric  Company,  Denver,  Colo. 
GEORGIA:  Atlanta William  Rawson  Collier, 

Georgia  Railway  &  Light  Company,  Atlanta,  Ga. 
MINNESOTA:    Minneapolis G.  D.  Shepardson, 

University  of  Minnesota,  Minneapolis,  Minn. 
St.  Paul A.  L.  Abbott, 

185  East  4th  Street,  St.  Paul,  Minn. 
OHIO:  Cleveland Ward  Harrison, 

Nela  Park,  Cleveland,  Ohio. 
WASHINGTON:  Seattle Fred.  A.  Osborn, 

University  of  Washington,  Seattle,  Wash. 


TRANSACTIONS 

OF    THE 

Illuminating  Engineering  Society 

Published  monthly,  except  during  July,  August,  and  September,  by  the 

ILLUMINATING    ENGINEERING   SOCIETY 

General  Offices:  29  West  Thirty-Ninth  Street.  New  York 


Vol.  VIM 


DECEMBER.  1913 


No.  9 


Council  Notes. 

A  regular  meeting  of  the  Council  was 
held  December  12,  1913,  in  the  general 
offices  of  the  society,  29  West  39th 
Street,  New  York.  In  attendance  were : 
C.  O.  Bond,  president;  J.  W.  Cowles, 
Ward  Harrison,  Joseph  D.  Israel,  gen- 
eral secretary;  V.  R.  Lansingh,  C.  A. 
Littlefield,  L.  B.  Marks,  treasurer; 
Preston  S.  Millar,  J.  Arnold  Norcross, 
C.  J.  Russell,  W.  J.  Serrill  and  G.  H. 
Stickney. 

The  meeting  was  called  to  order  at 
10:35  a.m.  by  President  Bond. 

To  supplement  the  sixth  paragraph  of 
the  minutes  of  the  October  Council 
meeting  (page  7  of  the  minutes  of  the 
present  administration)  the  following 
amendment  was  adopted : 

The  Illuminating  Engineering  Society  hereby 
expresses  its  adherence  and  support,  through 
the  United  States  National  Committee,  to  the 
International  Commission  on  Illumination;  and 
as  an  earnest  of  its  attitude  herewith  appro- 
priates its  assigned  quota  of  expenses  for  the 
ensuing  year,  $100.00,  to  the  use  of  the  said 
National  Committee. 

With  the  foregoing  amendment  the 
October  minutes  were  adopted. 

Mr.  Israel  reported  that  the  total 
membership  of  the  society  as  of  Decem- 
ber 10,  including  applications  and  resig- 
nations to  be  presented  at  the  meeting, 
was  1,401.  Counting  the  applications 
for  sustaining  membership  presented  at 
the  meeting  there  were  31  sustaining 
members,  which  are  not  included  in  the 


foregoing  membership  figure.  The  ex- 
penditures for  the  first  two  months  of 
the  present  fiscal  year  was  said  to  have 
totaled  $3,217.31. 

Vouchers  Nos.  1513  to  1548  inclusive, 
aggregating  $1,048.46,  were  authorized 
paid  subject  to  subsequent  approval  by 
the  Finance  Committee. 

In  accordance  with  recommendations 
contained  in  a  report  from  the  Finance 
Committee  it  was  voted  (1)  to  transfer 
the  account  of  the  society  now  in  the 
Lincoln  National  Bank  to  the  Central 
Trust  Company  at  Madison  Avenue  and 
42nd  Street,  New  York,  N.  Y.,  provided 
interest  at  2^  per  cent,  on  the  bank 
balance  cannot  be  obtained  from  the 
Lincoln  National  Bank;  the  Central 
Trust  Company  has  agreed  to  pay  the 
society  interest  at  that  rate  on  balances 
of  not  less  than  $1,000  and  not  more 
than  $10,000;  (2)  to  increase  the 
monthly  salary  of  Miss  Claire  Goldblatt, 
an  assistant  in  the  office  of  the  society, 
from  $44  to  $54,  and  (3)  to  authorize 
the  printing  of  another  edition  of  10,000 
copies  of  the  illumination  primer, 
"Light:  Its  Use  and  Misuse." 

Verbal  reports  of  progress  were  re- 
ceived from  the  following  committees : 
Sustaining  Membership,  Education,  and 
Section  Development.  The  reports 
were  accepted  with  commentation. 

Mr.  C.  A.  Littlefield,  chairman  of  the 
1913  Convention  Committee,  presented 
the  final  report  of  his  committee.     The 


TRANSACTIONS    I.    E.    S.  —  PART 


receipts  and  disbursements  were  said  to 
have  amounted  to  $1,875  and  $1,705.08 
respectively.  Accompanying  the  report 
was  a  scrap  book  outlining  the  way  the 
convention  was  conducted  and  including 
samples  of  various  letters  and  literature 
which  had  been  issued. 

Whereupon  it  was  resolved  to  extend 
to  the  committee  a  very  hearty  vote  of 
thanks  for  the  able  and  successful  man- 
ner in  which  the  1913  convention— prob- 
ably the  best  convention  ever  held  by 
the  society — was  conducted. 

Resolved,  that  a  special  vote  of  thanks 
be  transmitted  to  the  Papers  Committee 
of  the  previous  administration  for  hav- 
ing arranged  for  the  1913  convention 
an  excellent  program  of  papers,  which 
has  been  pronounced  the  best  balanced 
set  of  papers  ever  presented  before  a 
meeting  of  the  society. 

A  vote  of  thanks  was  extended  to 
Mr.  G.  H.  Stickney  for  his  able  presen- 
tation of  a  lecture  on  industrial  lighting 
at  a  session  of  the  International  Expo- 
sition of  Safety  and  Sanitation  held 
under  the  auspices  of  the  American 
Museum  of  Safety  in  the  Grand  Cen- 
tral Palace. 

Mr.  G.  H.  Stickney  reported  that  the 
contributions  referred  to  in  the  fore- 
going paragraph  had  been  sent  to  the 
office  of  the  society  and  that  he  pro- 
posed to  submit  bills  for  installing  the 
exhibit  to  be  paid  out  of  this  fund.  This 
procedure  was  accepted  and  accordingly 
the  Council  voted  to  authorize  payment 
of  such  bills  up  to  $300,  the  amount  of 
the  fund. 

It  was  resolved  that  a  vote  of  thanks 
be  sent  to  the  members  of  the  Lighting 
Exhibit  Committee  for  their  excellent 
services  in  arranging  the  foregoing  ex- 
hibit, and  to  the  Electrical  Testing 
Laboratories  for  its  kind  assistance  in 
the  work. 

Reports  on  section  activities  were  re- 


ceived from  the  following  vice-presi- 
dents: J.  W.  Cowles,  New  England; 
G.  H.  Stickney,  New  York;  W.  J  Ser- 
rill,  Philadelphia;  Ward  Harrison 
Pittsburgh;  and  J.  R.  Cravath,  Chicago! 
Mr.  G.  H.  Stickney  reported  infor- 
mally on  the  work  of  the  Papers  Com- 
mittee of  which  he  is  chairman. 

The  following  appointments  to  com- 
mittees were  confirmed: 

Nomenclature  and  Standards:  C.  H 
Sharp,  Louis  Bell,  C.  O.  Bond,  E.  P. 
Hyde,  H.  E.  Ives,  L.  B.  Marks',  A  S 
McAllister,  E.  B.  Rosa;  advisory  mem- 
bers: Andre  Blondel,  Hans  Bunte 
Vivian  B.  Lewes  and  C.  C.  Paterson. 

Section  Development:  Alan  Bright, 
C.  M.  Cole,  L.  B.  Eichengreen,  J.  B.' 
Jackson,  C.  L.  Law. 

Advertising:  B.  F.  Fisher,  Jr.,  F.  H. 
Gale,  J.  C.  McQuiston  and  Robert  F. 
Pierce. 

Popular  Lectures  :    Sub-Committee  on 
Residence    Lighting:      E.    J.    Edwards, 
chairman;  C.  R.  Clifford,  S.  G.  Hibben, 
J.  W.  Lee  and  J.  L.  Wiltse;  Sub-Com- 
mittee on  Store  Lighting:   A.  L.  Powell 
chairman;    J.    M.    Coles,    F.    H.    Gilpin,' 
D.  A.  Bowen,  C.  B.  Graves  and  A.  S.' 
Ives;      Sub-Committee     on     Industrial 
Lighting:  G  C.  Keech,  chairman;  Ward 
Harrison,  C.  E.  Stephens,  A.  J.  Sweet 
M.   H.   Flexner,   C.  A.   Luther  and   H.' 
W.   Shalling;   Sub-Committee  on   Office 
Lighting:      C.    E.     Clewell,    chairman; 
C.    L.    Law,    J.    G.    Henninger,    T.    W. 
Scofield. 

Papers :  Theodore  H.  Piser  and  Pro- 
fessor W.  E.  Wickenden. 

The  resignation  of  Albert  Jackson 
Marshall  as  a  director  of  the  society 
was  accepted.  It  was  voted  to  extend 
Mr.  Marshall  a  vote  of  thanks  for  his 
past  services  to  the  society. 

Mr.  J.  Arnold  Norcross,  secretary  and 
treasurer  of  the  New  Haven  Gas  Light 


TRANSACTIONS    I.    E.    S.  —  PART    I 


Company,  80  Crown  Street,  New  Haven, 
Conn.,  was  elected  a  director  of  the 
society  to  fulfill  the  unexpired  term  of 
the  vacancy  caused  by  the  election  of 
Mr.  C.  O.  Bond  to  the  presidency,  viz., 
December  12,  1913,  to  September  30, 
191S. 

Mr.  Alten  S.  Miller,  vice-president  of 
Humphreys  &  Miller,  Inc.,  165  Broad- 
way, New  York,  N.  Y.,  was  elected  a 
director  to  fulfill  the  unexpired  term  of 
Mr.  Albert  Jackson  Marshall,  resigned, 
vis.,  from  December  12,  1913,  to  Sep- 
tember 30,  1 914. 

Mention  was  made  of  a  letter  received 
from  Mr.  W.  F.  Durand,  executive  sec- 
retary of  the  International  Engineering 
Congress  which  is  to  be  held  in  San 
Francisco  in  1915.  This  letter  stated 
that  the  Congress  is  backed  by  the  five 
oldest  engineering  societies  and  that  the 
papers  to  be  presented  before  the  con- 
gress which  might  be  of  interest  to 
illuminating  engineers  would  be  included 
with  some  general  papers  in  a  miscella- 
neous volume,  the  contents  and  char- 
acter of  which  had  not  yet  been  deter- 
mined. 

Invitations  to  hold  the  1914  Conven- 
tion of  the  society  in  Cleveland  were 
received  from  local  civic,  professional 
and  commercial  organizations  and,  con- 
jointly, from  representatives  of  a  num- 
ber of  manufacturing  companies  and 
several  colleges. 

The  president  was  directed  to  appoint 
a  special  committee  to  consider  the  time 
and  place  of  the  1914  convention.  The 
aforementioned  invitations  were  referred 
to  that  committee. 

It  was  voted  to  hold  the  regular 
meetings  of  the  Council  during  the  rest 
of  the  present  administration  on  the 
afternoon  of  the  second  Thursday  of 
the  month  at  2  p.  m. 

The  meeting  adjourned  at  1  .25  p.  m. 


Section  Notes. 

CHICAGO    SECTION 

The  December  meeting  of  the  Chicago 
Section  was  held  on  the  10th  at  the 
residence  of  Mr.  A.  D.  Curtis  of  the 
National  X-Ray  Reflector  Co.  Mr. 
Curtis  read  a  paper  entitled  "Five  Years 
Progress  in  the  Indirect  Lighting  of  the 
Home."  A  number  of  different  types 
of  lighting  appliances  for  residence 
lighting  was  shown  in  connection  with 
the  paper.  Seventy-five  members  and 
guests  were  present. 

At  the  January  meeting,  which  will  be 
held  on  the  27th,  at  the  Art  Institute, 
Mr.  J.  B.  Jackson  will  read  a  paper 
entitled  "Planning  Lighting  Installa- 
tions." Mr.  W.  A.  Durgin  will  give  the 
second  one  of  his  20-minute  talks  on 
the  "The  Fundamentals  of  Illumination." 

NEW    ENGLAND    SECTION 

A  meeting  of  the  New  England  Sec- 
tion was  held  in  the  Auditorium  of  the 
Edison  Illuminating  Company  of  Bos- 
ton, December  9,  1913.  Mr.  L.  A.  Haw- 
kins of  the  Research  Laboratory  of  the 
General  Electric  Company  read  a  paper 
entitled  "The  Nitrogen-Filled  Lamp  and 
Its  Possibilities."  One  hundred  and 
forty  members  and  guests  were  present. 

NEW    YORK    SECTION 

A  regular  meeting  of  the  New  York 
Section  was  held  in  the  United  Engi- 
neering Societies  Building,  29  West  39th 
Street,  New  York,  December  11,  1913. 
Two  papers,  "Reasons  for  Styles  in 
Architecture"  and  "Periods  of  Archi- 
tecture as  Applied  to  Fixture  Design," 
were  presented  by  Messrs.  Frank  E. 
Wallace  and  Howard  E.  Watkins,  re- 
spectively. Both  papers  called  forth 
some  very  interesting  discussions.  One 
hundred  and  forty  members  and  40 
guests  were  present.     A  dinner  preced- 


TRANSACTIONS    I.    E.    S. — PART    I 


ing  the  meeting  at  Keens'  Chop  House 
in  West  36th  Street,  was  attended  by 
30  members  and  guests. 

PHILADELPHIA    SECTION 

The  Philadelphia  Section  held  a 
meeting  December  8  at  the  Engineers 
Club,  1317  Spruce  Street.  The  follow- 
ing papers  were  presented :  "Railway- 
Car  Lighting"  by  G.  H.  Hulse,  "The 
Measurement  of  Brightness  and  Its  Sig- 
nificance" by  Dr.  H.  E.  Ives,  and  "The 
Mercury  Quartz  Tube  Lamp"  by  Mr. 
M.  D.  Bucknam.  A  220-volt  quartz  tube 
lamp  was  demonstrated  by  Mr.  H.  Cal- 
vert. Mr.  J.  Stilwell  exhibited  some 
water  sterilizing  apparatus.  One  hun- 
dred and  sixty-five  members  and  guests 
attended  the  meeting.  At  a  dinner  at 
the  Engineers  Club  preceding  the  meet- 
ing 50  members  and  guests  were  present. 

The  program  of  meetings  and  papers 
for  the  rest  of  the  season  is  as  follows : 

Friday,  January  6. 
"Deficiencies  of  the  Method  of  Flicker 
for   the    Photometry   of    Lights    of 
Different  Colors." 

By  Prof.  C.  E.  Ferree. 

Saturday,  February  7. 
Meeting  under  the  Auspices  of  Drexel 

Institute. 
"Light  and  How  to  Use  It." 

By  Mr.  C.  O.  Bond,  President 
of  I.  E.  S. 

Wednesday,  February  18. 
Joint   Meeting  with   Franklin   Institute. 
"Artificial  Daylight." 

By  Dr.  Herbert  E.  Ives. 

Friday,  March  20. 
"Lighting    and    Signalling    Systems    of 
Subways." 

By  Air.  F.  D.  Bartlett. 
"The  Sun — The  Master  Lamp." 

By  Prof.  James  Barnes. 


Thursday,  April  9. 
J^int   Meeting  with   Franklin   Institute. 
"Recent    Developments    in    the    Art    of 
Illumination." 

By  Mr.  Preston  S.  Millar. 
Friday,  April  17. 
"The  Structure  of  the  Normal  Eye  and 
its  Ability  to  Protect  Itself  Against 
Ordinary  Light." 

By  Dr.  Wendell  Reber. 
"Glassware  for  Illumination  and  Other 
Purposes." 

By  Mr.  James  Gillinder. 

Friday,  May  15. 

Mass    Meeting   of    all    the    Engineering 

Societies  of  Philadelphia  and 

Vicinity. 

Special  Program  to  be  arranged  and  to 

include  an  address  on 
"The    Relation    of     Engineers    to    the 
Progress  of  Civilization." 

By  Dr.  Chas.  Proteus  Steinmetz. 

PITTSBURGH   SECTION 

The  November  meeting  was  held  on 
the  28th  in  Thaw  Hall  of  the  University 
of  Pittsburgh.  Dean  F.  L.  Bishop  of 
the  Engineering  School  of  the  latter 
university  gave  a  lecture  on  "The  Ptrys- 
ics  of  Lighting."  Fifty-three  members 
and  guests  were  present. 

On  December  12,  the  Pittsburgh  Sec- 
tion held  a  joint  session  with  the  local 
chapter  of  the  American  Institute  of 
Electrical  Engineers  in  the  Auditorium 
of  the  Engineers  Society  of  Western 
Pennsylvania.  One  hundred  and  thirty- 
two  members  and  guests  of  both  socie- 
ties were  present.  The  following  papers 
were  presented  :  "A  Problem  in  Boule- 
vard Lighting"  by  J.  M.  Froelich; 
"Variables  in  Street  Lighting"  by  J.  F. 
Martin;  and  "Some  Aspects  of  Free 
Lamp  Renewals"  by  T.  F.  Campbell. 

The  following  program  of  papers  and 
meetings  has  been  arranged : 


TRANSACTIONS   I.    E.    S.— PART   I 


January  i6th. 

(i)  "A  Photographic  Analysis  of 
Diffusing  Units  with  Varying  Indirect 
Component"  by  E.  B.  Rowe  and  H.  H. 
Magdsick  of  the  National  Electric  Lamp 
Association,  Cleveland,  Ohio.  The 
speakers  will  have  photometric  curves 
on  present  types  of  commercial  "semi- 
direct"  glassware  showing  the  effect 
or  variations  in  contour,  density  and 
lamp  arrangement  on  absorption  and 
transmission,  as  well  as  on  reflecting 
and  illuminating  efficiency. 

(2)  "The  Relation  of  the  Engineer  to 
the  Problems  of  Fixture  Design"  by 
A.  B.  Wilson  and  F.  J.  Blaschke  of  the 
National  Electric  Lamp  Association, 
Cleveland,  Ohio.  This  paper  deals  with 
the  object  of  a  lighting  installation,  ex- 
plains what  a  fixture  really  is  and  illus- 
trates its  correct  and  incorrect  applica- 
tion, mentions  co-operation  between 
engineer  and  fixture  maker,  states  how 
designers'  ideas  are  carried  out.  It 
also  shows  how  the  lighting  fixture  is 
an  element  in  the  advancement  of  the 
art  of  illumination. 

February   13TH. 

"Lighting  of  Railroad  Yards"  by  A. 
C.  Cotton  of  the  Pennsylvania  Railroad 
and  Harold  Kirschberg  of  the  Lighting 
Specialties  Company.  The  authors  will 
discuss  the  various  elements  entering 
into  the  lighting  of  track  scales  and 
classification  yards,  together  with  the 
difficulties  experienced  with  same,  and 
how  they  are  best  overcome. 
March  13TH. 

"Modern  Gas  Lighting"  by  S.  B. 
Stewart,  Contract  Agent  for  Consoli- 
dated Gas  Company,  Pittsburgh,  Pa. 
This  meeting  will  be  devoted  to  the  dis- 
cussion of  gas  arcs  as  applied  to  modern 
illuminating  systems.  A  number  of 
prominent  manufacturers  and  operators 
will  be  present  and  take  part  in  the  dis- 
cussion. 


April  17TH. 
"The  Development  of  Flame  Carbon 
Arc  Lamps"  by  C.  E.  Stephens,  West- 
inghouse  Electric  &  Mfg.  Company. 
The  author  will  trace  the  growth  and 
development  of  this  popular  form  of 
illuminant  from  its  inception  down  to  the 
present  time,  showing  how  the  difficul- 
ties first  experienced  have  been  over- 
come, and  its  application  to  various 
fields. 


New  Members. 

At  a  meeting  of  the  Council  held 
December  12,  1913,  the  following  28 
applicants  were  elected  members  of  the 
society : 

Anderson,  C.  E. 

Educational  Department,  General 
Electric  Company,   Harrison,   N.  J. 

Arbogast,  O.  J. 

Salesman,  Commonwealth  Edison 
Company,  120  West  Adams  Street, 
Chicago,  111. 

Barnitz,  Frank  R. 

Assistant  Secretary,  Consolidated 
Gas  Company,  124  East  15th  Street, 
New  York  City. 

Bertsche,  Fred. 

Commonwealth  Edison  Company, 
120  West  Adams  Street,  Chicago, 
111. 

Breuggemeyer,  A.  H. 

Lighting  Solicitor,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 

Crowley,  Frank  M. 

Lighting  Salesman,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 

Dawson,  James. 

Salesman,  Commonwealth  Edison 
Company,  120  West  Adams  Street, 
Chicago,  111. 


TRANSACTIONS   I.    E.    S. — PART   I 


Donnelly,  Jas.  E. 

Light  Salesman,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 

Griner,  Charles  A. 

Salesman,  Commonwealth  Edison 
Company,  120  West  Adams  Street, 
Chicago,  111. 

Hecker,  L.  M. 

Salesman,  Contract  Department, 
Commonwealth  Edison  Company, 
120  West  Adams  Street,  Chicago. 
111. 

Horan,  William  H. 

Salesman,  Commonwealth  Edison 
Company,  120  West  Adams  Street, 
Chicago,  111. 

Hyedahl,  Nick. 

Commonwealth  Edison  Company, 
120  West  Adams  Street,  Chicago, 
111. 

Kaeder,  F.  J. 

Salesman,  Commonwealth  Edison 
Company,  120  West  Adams  Street, 
Chicago,  111. 

Keys,  Harvey  A. 

Light  Salesman,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 

Lancaster,  Walter  B. 

Physician,  101  Newbury  Street, 
Boston,  Mass. 

Long,  Claude  P. 

Solicitor,  Contract  Dept.,  Common- 
wealth Edison  Company,  120  West 
Adams  Street,  Chicago,  111. 

Lorenz,  J.  M. 

Salesman,  Central  Electric  Com- 
pany, 320  5th  Avenue,  Chicago,  111. 

Myers,  Romaine  W. 

Consulting  Engineer,  204  Bacon 
Building,  Oakland,  Cal. 

O'Brien,  John  C. 

Lighting  Salesman,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 


Peck,  Edward  L. 

Inspector,  Electrical  Testing  Labo- 
ratories, 80th  Street  and  East  End 
Avenue,  New  York  City. 

Prabel,  Frederick  C. 

Lighting  Salesman,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 

Reill,  Wilfred  J. 

Lighting  Salesman,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 

Reinach,  Hugo  B. 

Asst.  Gen'l.  Superintendent,  Con- 
solidated Gas  Company,  124  East 
15th  Street,  New  York  City. 

Rosenberg,  Maurice. 

General  Manager,  Shapiro  &  Aron- 
son,  20  Warren  Street,  New  York 
City. 

Rusch,  William  T. 

Assistant  to  Engineer  of  Utilization, 
Consolidated  Gas  Company,  124 
East  15th  Street,  New  York  City. 

Severn,  George  K. 

Salesman,  Commonwealth  Edison 
Company,  120  West  Adams  Street, 
Chicago,  111. 

Stilwell,  John. 

Efficiency  Engineer,  Consolidated 
Gas  Company,  124  East  15th  Street, 
New  York  City. 

Winans,  R.  K. 

Lighting  Agent,  Commonwealth 
Edison  Company,  120  West  Adams 
Street,  Chicago,  111. 


Additional  Sustaining  Members. 

The    following    five    companies    were 
elected     sustaining     members     of     the 
society,  December  12,  1913 : 
Consolidated   Gas    Company    of    New 
York. 
124    East    15th    Street,    New    York 
City. 


TRANSACTIONS    I.    E.    S. — PART    I 


National  X-Ray  Reflector  Company. 
235  West  Jackson  Boulevard,  Chi- 
cago, 111. 

Official    Representative :      Augustus 
D.  Curtis. 

Public  Service  Company  of  Northern 
Illinois. 
157  South  LaSalle  Street,  Chicago, 
111. 

Stone  &  Webster. 

147  Milk  Street,  Boston,  Mass. 

The  Cleveland  Electric  Illuminating 
Company. 
Cleveland,  Ohio. 

Official      Representative:        R.      H. 
Wright. 


International  Electrical  Congress. 

The  International  Electrical  Congress 
is  to  be  held  at  San  Francisco,  Septem- 
ber 13  to  18,  1915,  under  the  auspices 
of  the  American  Institute  of  Electrical 
Engineers  by  authority  of  the  Inter- 
national Electrotechnical  Commission, 
and  during  the  Panama-Pacific  Inter- 
national Exposition.  Dr.  C.  P.  Stein- 
metz  has  accepted  the  honorary  presi- 
dency of  the  Congress.  The  delibera- 
tions of  the  Congress  will  be  divided 
among  twelve  sections  which  will  deal 
exclusively  with  electricity  and  electrical 
practise.  There  will  probably  be  about 
250  papers.  The  first  membership  invi- 
tations will  be  issued  in  February  or 
March,  1914. 

Attention  is  drawn  to  the  distinction 
between  this  Electrical  Congress  and 
the  International  Engineering  Congress 
which  will  be  held  at  San  Francisco 
during  the  week  immediately  following 


the  Electrical  Congress.  The  Engineer- 
ing Congress  is  supported  by  the  Socie- 
ties of  Civil,  Mechanical  and  Marine 
Engineers  and  by  the  Institutes  of  Min- 
ing and  Electrical  Engineers,  as  well  as 
by  prominent  Pacific  Coast  engineers 
who  are  actively  engaged  in  organizing 
it.  This  Congress  will  deal  with  engi- 
neering in  a  general  sense,  electrical 
engineering  subjects  being  limited  to 
one  of  the  eleven  sections  which  will 
include  about  twelve  papers,  treating 
more  particularly  applications  of  elec- 
tricity in  engineering  work. 

The  meeting  of  the  International 
Electrotechnical  Commission  will  be 
held  during  the  week  preceding  that  of 
the  Electrical  Congress. 


Nomenclature. 

At  the  next  meeting  of  the  Committee 
on  Nomenclature  and  Standards,  which 
will  take  place  in  February,  the  follow- 
ing definitions  will  come  up  for  dis- 
cussion : 

A  direct  lighting  system  is  one  in 
which  substantially  all  the  useful  light 
flux  comes  directly  from  the  illuminant, 
including  its  diffusing  or  reflecting 
auxiliaries. 

An  indirect  lighting  system  is  one  in 
which  substantially  all  the  useful  light 
flux  is  received  by  diffuse  reflection 
from  the  ceiling,  walls  or  other  diffusely 
reflecting  surfaces  of  relatively  large 
extent. 

Criticisms  and  comments  on  these 
definitions  may  be  sent  to  the  secretary 
of  the  committee,  Dr.  Clayton  H.  Sharp, 
80th  Street  and  East  End  Avenue,  New 
York,  N.  Y. 


TRANSACTIONS   I.    E.    S. — PART  I 

FINANCIAL  REPORT  FOR  FISCAL  YEAR 
ENDING  SEPTEMBER  30,   1913. 


Wm.  J.  Struss  &  Co. 

Certified  Public  Accountants 

93-99  Nassau  Street 

New  York 

October  24th,  19 13. 
Mr.  William  Cullen  Morris. 
Chairman,  Finance  Committee, 

Illuminating  Engineering  Society, 
29  West  39th  Street, 
New  York,  N.  Y. 
Dear  Sir:— 

In  accordance  with  your  instructions  we  have  examined  the 
books  and  accounts  of  the  Illuminating  Engineering  Society  for 
the  nine  (9)  months  ended  September  30th,  19 13. 

The  results  of  this  examination  are  set  forth  in  the  two  ex- 
hibits, attached  hereto,  as  follows  : 

Exhibit  "A  "—Balance  Sheet,  September  30th,  1913. 

Exhibit  "B" — Earnings  and  Expenses,   for  the  nine 

months  ended  September  30th,  1913. 

We  hereby  certify  that  the  accompanying  Balance  Sheet  is  a 

true  exhibit  of  its  financial  condition  as  of  September  30th,  1913, 

and  that  the  attached  statement  of  Earnings  and  Expenses  is 

correct. 

Respectfully  submitted, 

Wm.  J.  Struss  &  Co., 

Certified  Public  Accountants. 

Illuminating  Engineering  Society  Balance  Sheet 

September  30TH,  1913. 

Exhibit  "A" 

Assets 

Cash  on  hand  and  in  bank $2,443. 53 

Accounts  receivable — 

191 2  Miscellaneous $      20.28 

1913  Miscellaneous 279.40 

1913  Dues 430.00 

1913  Initiation  fees 120.00 

1913  Advertising  • 4.46 

i9i4Dues-. 120.00 

1914  Dues — sustaining  members 675.00 

1,649.14 


TRANSACTIONS   I.    E.    S. — PART   I  9 

Investments — 

Northern    Pacific  &  Great   Northern 

Railway    Bonds — $2,000.00     (book 

value) 1,920.00 

Furniture  and  fixtures 894.71 

Less  depreciation  15  per  cent 134.20  760.51 

Badges  on  hand 45-oo 

Accrued  interest  on  bonds 20.00 

$6,838.18 
Liabilities. 
Advanced  and  unearned  items — 

Advance  Dues,  1913 $    250.00 

Fees,  1913 3500 

Advertising,  1913 4.93 

Dues,  1914 160.00 

Fees,  1914 20.00 

Dues    sustaining    members 

1914 60.00 

Unearned  Sustaining  members  dues  935-42 

Dues...    •    1,717.50      3,182.85 

Outstanding  debts  (estimated) 1 ,789.93 

Surplus  Balance,  January  2,  1913 1,453-94 

Back  dues  collected  1912 73.00 

Convention  fund  surplus  1912.-.  177-74 

Miscellaneous 79-88 

Net  gain  for  the  nine  months, 
ended  September  30,  1913,  as 
per  Exhibit  "B" 80.84       1,865.40 

$6,838.18 

Illuminating  Engineering  Society 

Statement  of  Earnings  and  Expenses  for  the  Nine  Months  Ended 

September  30TH,  1913. 

Exhibit  "B" 

Earnings. 

Members  dues I5.207.50 

Initiation  fees 360.00 

Advertising 1,097. 14 

Profit  on  badges  sold 6.50 

Certificates 1.70 

Interest  on  bonds 60.00 

Miscellaneous  sales  of  Transactions  •  •  •  211.52 

Illumination  Primer 616.25 

Sustaining  members  dues 689.58 

Annual  meeting 6.30 

$8,256.49 


10  TRANSACTIONS    I.    E.    S. — PART    I 

Expenses. 

Transactions $1,844.72 

General  office — 

Salaries $2,226.14 

Rent 633.15 

Postage 236.35 

Telephone  and  telegraph  • .  •  •  143 .96 

Printing  and  stationery 294.24 

Miscellaneous 235.00      3,768.84 

New  York  Section 245.03 

Chicago  Section 217.04 

New  England  Section 53 .78 

Pittsburgh  Section 149-73 

Philadelphia  Section 130.88 

1913  Convention 257.67 

1913  Election  expense 57-44 

Committee  on — 

Nomenclature  and  Standards  14-67 

Glare 60.00 

Research 17-5° 

Popular  Lectures  4.25 

Collegiate  Education 10.75 

Papers 14.35 

Sustaining  Membership 6.40 

Joint     meetings     with     other 

societies 43-57 

Depreciation  on  furniture  and 

fixtures 134. 20 

Miscellaneous  expense 100.65 

Advance  copies  and  reprints  •  •  9.93 

Exchange  and  discount 9.25 

Unpaid  accounts  (estimated) — 

General  office 30.00 

Transactions 5.00 

Miscellaneous 75 .00 

1913  Convention 915.00       1,025.00      8,175.65 


Excess  of  earnings  over   ex- 
penses   $    80. S4 


■r- 


6'VH 


TRANSACTIONS 


OF  THE 


Illuminating 
Engineering  Society 

DECEMBER,  1913 
PART  II 


Papers,  Discussions  and  Reports 


[  DECEMBER,  1913  ] 

CONTENTS  -  PART  II 


The  Lighting  of  Show  Windows.     By  H.  B.  Wheeler 555 

Modern  Practise  in  Street  Railway  Illumination.     By  S.  G. 
Hibben 589 

Church  Lighting.     By  Robert  B.  Ely 613 

Some  Studies  in   Accuracy   of   Photometry.      By   Evan  J. 
Edwards  and  Ward  Harrison 633 

The  Status  of  the  Lighting  Art 652 

Annual  Report  of  General  Secretary.     By  Joseph  D.  Israel  683 


**■>'• 


THE  LIGHTING  OF  SHOW  WINDOWS.* 


BY  H.  B.  WHEELER. 


Synopsis:  The  lighting  of  a  number  of  typical  show  windows  is  dis- 
cussed with  respect  to  the  intensity  required  along  the  line  of  trim,  and 
the  selection  of  reflectors,  their  spacings  and  methods  of  installation. 
Illustrations,  plans  and  illuminometer  readings  and  curves  of  a  number 
of  well  lighted  windows  are  given.  The  material  is  arranged  so  as  to 
make  the  selection,  spacing,  and  installation  of  reflectors  very  simple. 
The  paper  is  divided  into  two  sections :  The  first  covers  methods  of  il- 
luminating various  typical  windows.  The  second  gives  test  data  on  the 
same. 

Up  to  the  present  time  very  little  data  has  been  presented  on 
the  lighting  of  show  windows.  That  which  has  been  given 
pertains  chiefly  to  the  requirements  of  good  show  window  light- 
ing, the  influence  of  surroundings  on  the  treatment  of  a  window 
and,  occasionally,  a  description  of  some  one  notable  installation. 
I  have  assumed  that  the  foregoing  matters  are  well  established 
facts,  and  therefore  this  paper  is  devoted  to  methods  of  obtaining 
these  good  results  in  window  illumination. 

Typical  show  windows  have  been  grouped  in  three  classes 
which  for  convenience  are  designated  as  A,  B  and  C;  and  the 
more  special  windows  in  two  other  classes,  referred  to  as  D  and 
E-  For  each  class  I  show  a  window  of  the  type  under  considera- 
tion, the  reflector  which  was  used,  the  method  of  installing  the 
reflector,  and  a  distribution  curve  of  the  reflector. 

The  questions  of  selecting  the  proper  reflector  for  a  window, 
and  the  spacings  to  be  used,  to  give  the  desired  intensity  of  illum- 
ination along  the  line  of  trim,  are  discussed,  and  I  have  added 
recommendations  of  the  intensities  I  have  found  good  practise 
for  show  windows.  Lamps,  shade  holders,  window  drapes  (val- 
ances), and  other  features  of  window  trimming  that  are  of  vital 
interest  to  the  illuminating  engineer  are  discussed  briefly.  The 
paper  also  includes  the  results  of  tests  made  on  typical  windows 
of  the  various  classes  discussed. 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not   responsible  for  the  statements  or 
opinions  advanced  by  contributors. 

2 


556  TRANSACTIONS  I.  E.   S. — PART  II 

In  general  there  are  two  types  of  windows  :  ( i )  open  back ; 
(2)  boxed-in.  In  the  first  class  the  display  is  usually  very  low, 
in  most  cases  being  practically  horizontal.  The  problem  here  is 
merely  to  provide  the  proper  intensity  of  horizontal  illumination. 
Since  these  windows  are  generally  quite  shallow  and  high,  they 
require  a  concentrating  type  of  reflector  for  the  very  shallow 
window,  and  a  semi-concentrating  for  the  deeper  ones. 

The  reflectors  shown  in  Figs.  10  and  7,  Nos.  755  and  780, 
respectively,  are  most  suitable  for  such  windows. 

The  following  classification  I  believe  takes  into  account  all 
of  the  various  types  of  boxed-in  windows  encountered  to-day. 

TYPICAL  WINDOWS. 

Class  A  Windows :  Height  equal  to  depth.  High  trim.  Aver- 
age height  9  ft.  (2.74  m.).  Depth  9  ft.  (2.74  m.).  Height  of 
trim  9  ft.  (2.74  m.). 

Class  B  Windows:  (1)  Height  1^  times  depth.  Medium 
trim.  Average  height  10  ft.  (3.04  m.).  Depth  6  ft.  (1.83  m.). 
Height  of  trim  7  ft.  (2.13  m.). 

(2)  Height  1  J/2  times  depth.  High  trim.  Average  height 
10  ft.  (3.04  m.).  Depth  6  ft.  (1.183  m-)-  Height  of  trim  9  ft. 
(2.74  m.) 

Class  C  Windows :  Height  2  times  depth.  Medium  trim. 
Average  height  12  ft.  (3.65  m.).  Depth  6  ft.  (1.83  m.).  Height 
of  trim  7  ft.  (2.13  m.). 

SPECIAL  WINDOWS. 

Class  D  Windows :  Height  equal  to  depth.  High  trim.  Aver- 
age height  5  ft.  (1.52  m.).  Depth  6  ft.  (1.83  m.).  Height  of  trim 
5  ft.  (1.52  m.). 

Class  E  Windows :  Height  2  times  depth.  Low  trim. 
Average  height  5  ft.  (1.52  m.).  Depth  2  ft.  6  in.  (0.76  m.). 
Height  of  trim  2  ft.  (0.61  m.). 

The  height  of  the  window  is  always  measured  from  the  floor 
to  the  ceiling;  the  depth,  from  the  glass  front  to  the  back;  the 
trim,  from  the  floor  up. 

Class  A   Windozvs. — In  Fig.   1  is  shown  a  Class  A  window. 


whkklkk:   the  lighting  of  show  windows  557 

Windows  of  this  class  usually  are  trimmed  up  high  on  the  back- 
ground, and  hence  require  a  reflector  which  distributes  the  light 
flux  over  the  angle  zero  to  90  degrees. 

The  reflector  shown  in  Fig.  2  has  been  designed  to  meet  this 
condition ;  its  distribution  curve  is  shown  in  Fig.  4.  This  re- 
flector is  a  non-symmetrical  reflector,  with  a  portion  of  the 
front  cut  away,  to  permit  the  light  flux  to  escape  horizontally. 
Hence  a  large  portion  of  the  lamp  is  exposed  to  view  from  within 
the  window.  Mirrors  should  not  be  placed  in  the  upper  part  of 
the  background,  in  order  to  avoid  the  possibility  of  seeing  an 
image  of  the  lamp  when  observing  the  window  from  the  street. 

The  practise  of  using  mirrors  in  any  window  should  be  dis- 
couraged as  images  of  surrounding  objects  are  generally  present, 
which  detract  from  the  goods  on  display.  One  of  the  chief 
reasons  why  merchants  desire  mirrors,  is  the  fact  that  they  be- 
lieve the  observer  will  be  enabled  to  see  both  the  front  and  back 
of  the  objects  in  the  window.  However,  the  brightness  to  which 
the  back  of  the  objects  is  illuminated  is  so  low  that  the  results 
are  not  very  satisfactory.  In  addition  to  this,  much  brighter 
images  of  surrounding  objects  serve  to  detract  rather  than  add 
to  the  effect  soueht. 


Fig-  1.— A  Class  A  window  ;  height  9  ft.,  depth  9  ft.,  equipped  with  60-watt  clear 
tungsten  lamps  in  reflectors  (Fig.  2)  on  15-inch  centers. 


553 


TRANSACTIONS  I.   E.   S. — PART  II 


Fig.  2.— Reflector  (No.  777). 
Details  for  installing  in 
windows  shown  in  Fig.  3. 


A/o/e  -    With  unskirted    lamps    use  extension  socket, 

Fig-  3-— Details  for  installing  in  windows  reflector  shown  in  Fig.  2. 


wheeler:   the  lighting  oe  show  windows 


i59 


Z 
< 

CANDLE.  POWER 

<x> 
z 
< 

CANDLE    POWER 

REFLECTOR  Ho  777 
HOLDER  24   Form  '0' 
LAMP  Clear  Tungsten  S-2*i  -A 
WATTS  636          VOLTS  »5S 

CURVE   A 

CURVE B 

CURVE    A 

CURVE  8 

R 

L 

R 

R 

u 

R 

0 

166.3 

166-3 

166-3 

90 

151.6 

5' 

£65.2 

110-9 

159.  S 

95 

IIO.Z 

15 

nz.4 

385 

HZ-S 

105 

ZS.7 

Q 

231.0 

193 

112.7 

115 

I.Z 

35 

ZJ1.9 

63 

34.  Z 

125 

40 

135 

45 

20*.  9 

71.7 

K5 

55 

239.7 

S8.2 

155 

65 

Z6S.6 

39.6 

165 

75 

2413 

19.7 

175 

85 

1916 

3.4 

ISO 

Fig.  4- 


-Photometric  distribution  curve  and  data  on  reflector  No. 
Fig.  2,  with  1.25-watt-per-candle  tungsten  lamp. 


777,  shown  in 


560 


TRANSACTIONS  I.  £.  S. — PART  II 


u 
o 

z 
< 

CANDLE   POWER 

UI 

z 

CANDLE  POWER 

REFLECTOR     No-780 
HQLPER   314'  Form  "A" 
lamp  Clear  Tungsten  SSO'A 
WATTS  WOO     VOLTS.  UO.O 

CURVE  A 

CURVES 

CURVE  A 

CURVE  B 

R 

L 

R 

R 

L 

R 

0 

4/5.0 

415.0 

4IS.0 

90 

5 

388.0 

411-0 

408.0 

95 

15 

301.0 

Z47.0 

408. 0 

105 

IS 

307.0 

116.0 

400.0 

1)5 

35 

380.0 

tt.O 

Z97-0 

its 

40 

135 

45 

3S3.0 

4.Z 

137.0 

W5 

55 

156.0 

246 

155 

65 

BO.S 

3.1 

165 

75 

2.4 

175 

85 

185 

Fig.  5- — Photometric  distribution  curve  and  data  on  reflector  (No.  780)  shown 
in  Fig.  7,  with  1.08-watt-per-candle  clear  tungsten  lamp. 


wheeler:   the  lighting  of  show  windows 


561 


Class  B  Windows. — In  Fig.  6  is  shown  a  Class  B  window. 
The  more  common  windows  of  this  class  are  trimmed  only  to  a 
medium  height,  but  occasionally  some  are  found  in  which  the 
trim  is  carried  up  high.  The  first  sub-division  requires  the  use 
of  a  reflector  which  distributes  the  light  flux  in  the  angle  zero 
to  55   degrees. 

The  reflector  shown  in  Fig.  7  has  been  designed  to  accomplish 
this  result.     Its  distribution  curve  is  shown  in  Fig.  5. 

Reflector  No.  780  not  only  confines  the  light  flux  where  de- 
sired, but  hides  the  lamp  filament  from  view  within  the  store, 
with  even  a  medium  height  background.  Curtains  above  the 
background  are  often  used  to  conceal  the  reflectors,  (see  Fig. 
6.)    ' 

For  the  second  sub-division  it  is  necessary  to  use  a  combina- 
tion of  the  reflectors  shown  in  Figs.  2  and  7  in  order  to  get  the 
desired  results. 

A  solid  background  is  advisable  with  this  latter  arrangement 
unless  suitable  curtains  are  used. 


Fig.  6.— A  Class  B  window  ;  height  12  ft.,  depth  7  ft.,  background  S  ft.,  equipped 
with  clear  100-watt  tungsten  lamps  in  reflectors  (Fig.  7)  on  12-iuch  centers. 


562 


TRANSACTIONS  I.  E.   S. — PART  II 


Fig.  7.— Reflector  (No.  780).     Details  for  installing  in  windows  shown  in  Fig.  8. 


Fig.  8.— Details  for  installing  in  windows  reflector  (No.  780)  shown  in  Fig^-7. 


wheeler:   the  lighting  oe  show  windows 


563 


Class  C  Windows. — In  Fig.  9  is  shown  a  Class  C  window. 
In  this  class  of  window  the  trim  is  carried  up  to  only  a  medium 
height.  A  reflector  which  concentrates  the  light  flux  in  the 
angle  intercepted  by  the  line  of  trim,  most  nearly  meets  the  con- 
ditions. This  angle  is  relatively  smaller  than  for  most  other 
types  of  windows,  and  hence  presents  a  much  harder  problem  to 
be  dealt  with.  It  is  not  only  necessary  to  get  high  concentration 
on  the  floor  of  the  window,  but  the  background  also  must  be 
properly  illuminated  to  the  top  of  the  trim,  and  the  light  flux 
cut-off  quite  sharply  at  this  point. 

Fig.  10  shows  a  reflector  which  has  been  designed  to  give  the 
required  results.     Its  light  distribution  curve  is  shown  in  Fig.  12 


Fi&-  9-— A  Class  C  window,  height  16  ft.  6  in.,  depth  8  ft.,  background  8  ft.,  equipped 
with  ioo- watt  clear  tungsten  lamps  in  reflectqrs  (Fig.  io)  on  i8-inch  centers. 


564 


TRANSACTIONS   I.   E.   S. PART  II 


Fig.  io.— Reflector  (No.  755)  details  for  installing  in  windows  shown  in  Fig.  11. 


Fig.  11. — Details  for  installing  in  windows  reflector  (No.  755)  shown  in  Fig.  10. 


WHEELER  :    THE  LIGHTING  OF  SHOW  WINDOWS  565 


IP 

z 

CANDLE  POWER 

Ui 

_i 

z 
< 

CANDLE  POWER 

1 
REFLECTOR    4*  7& 

holder  ,#  Form'A 

L^MP  Tungsten  S-JO-AC/ecr 

WATTS  1000        VOLTS  llOO 

CURVE  A 

CURVE  8 

CURVE   A 

CURVE  B 

R 

L 

R 

R 

L 

R 

0 

926.0 

9Z6.0 

926.0 

<» 

5 

895.0 

677.0 

931.0 

95 

15 

617.0 

66S.0 

7S7.0 

105 

15 

497.0 

z».o 

fd&O 

115 

35 

mu> 

70.0 

130,0 

IZS 

40 

135 

45 

2800 

6-6 

S.'.B 

MS 

55 

IIZ.0 

to 

164 

155 

65 

liS 

7.1 

165 

75 

17 

3J 

175 

85 

180 

Fig.  12.— Photometric  distribution  curve  and  data  on  reflector  (No.  755) 
shown  in  Fig.  10,  with  a  1.08-watt-per-candle  lamp. 


566 


TRANSACTIONS  I.  E.   S. — PART  II 


1 

CANDLE"  POWER 

ex. 

CANDLE  POWER 

REFLECTOR  M-  7S0 
HOLDER  Special  attached 
LAMP  Clear  Tungsten  J-  Zl- E 
WATTS  IZO-0        VOLTS  l'0-O 

CURVE  A 

CURVEB 

CURVE  A 

CURVE  B 

R 

L 

R 

R 

L 

R 

0 

4ZBM 

428.0 

4ZS.0 

90 

s 

fJZ.O 

413.0 

t28.0 

95 

15 

448.0 

330J/ 

404,0 

105 

25 

430M 

ZZ3.0 

354.0 

115 

J5 

401.0 

131.0 

zaz.o 

125 

40 

135 

45 

356.0 

6tt 

116.0 

145 

55 

314.0 

Z4.0 

116.0 

155 

65 

Z44.0 

6ZJ) 

165 

75 

96.6 

3&S 

175 

es 

IZ.I 

1S.7 

180 

-Photometric  distribution  curve  and  data  on  reflector  (No.  750)  shown 
in  Fig.  15,  with  1.12-watt-per-candle  clear  tungsten  lamp. 


wheeler:   the  lighting  of  show  windows  567 

Class  D  Windozvs. — This  class  of  window  is  found  chiefly  in 
cases  where  the  show  window  is  divided  into  two  tiers,  and  in 
shops  with  low  head  room,  situated  on  the  ground  floor. 
The  light  distribution  required  is  essentially  that  of  the  reflector 
shown  in  Fig.  2,  but  there  is  not  sufficient  head  room  in  these 
windows  to  permit  its  use. 

Fig.  15  shows  a  reflector  which  has  been  designed  to  meet  the 
latter  condition.  Its  light  distribution  curve  is  shown  in  Fig. 
13.  This  reflector  is  provided  with  an  adjustable  holder  as 
shown  in  Fig.  16,  and  the  lamps  are  placed  horizontally  in  the 
reflector.  It  is  usually  installed  with  the  flat  side  tilted  at  an 
angle  of  approximately  15  degrees  with  the  horizontal.  A  mir- 
ror background  should  not  be  used  in  a  window  of  this  type  for 
obvious  reasons. 

Class  E  Windozvs. — In  Fig.  18  is  shown  a  class  E  window. 
This  class  of  window  is  found  largely  in  jewelry  stores,  cigar 
stores,  and  shoe  stores.  The  line  of  trim  is  low,  and  frequently 
it  is  practically  flat.  This  window  requires  a  reflector  having 
a  light  distribution  curve  similar  to  that  of  the  reflector  shown  in 
Fig.  10.  This  reflector  of  course  is  too  large  for  a  window  of 
this  size. 

The  reflector  (Fig.  19)  installed  as  shown  in  Fig.  20,  has  been 
largely  used  for  this  type  of  window.  Its  distribution  curve  is 
shown  in  Fig.  14.  A  window  may  be  illuminated  in  this  man- 
ner very  satisfactorily,  but  since  the  reflector  is  symmetrical  it  is 
not  so  economical,  inasmuch  as  a  great  deal  of  the  light  flux  es- 
capes to  the  street  and  upper  portion  of  the  window.  The 
percentage  of  the  total  light  flux  incident  on  the  surface  of  the 
line  of  trim  is  low. 

At  the  present  time  a  small  non-symmetrical  reflector  like  that 
shown  in  Fig.  10  is  being  developed  for  this  class  of  windows. 

Preliminary  tests  on  this  new  reflector  with  an  unskirted  base 
60-watt  clear  bulb  tungsten  lamp  show  a  highly  concentrated 
light  distribution  very  similar  to  that  of  Fig.  12.  The  cut-off 
at  the  window  side  of  the  reflector  is  slightly  sharper  than  for 
the  No.  755  reflector. 


568 


TRANSACTIONS  I.  E.   S. — PART  II 


4*1 

jr 
i. 

CANDLE  POWER 

< 

CANDLE  POWER 

LAMP  AND  REFLECTOR 

BARE  LAMP 

LAMP 
&RER- 

BARE 
LAMP 

LAMP 
SrREFL. 

BARE 
LAMP 

ZONE" 

0-60 

0-90 

0-190 

0-60. 

0-90 

0-80 

LUMENS 

Wft? 

£Z£7 

■i"^? 

SZf.O 

0 

306.0 

90 

REFLECTOR  M.m           H0LPER#'^/ot  0 
LAMP  Tungsten    WATTS  6°-°      vo\js-M-0 

5 

ViSJ) 

95 

IS 

234.0 

105 

IS 

175.0 

115 

35 

MO.O 

\15 

40 

135 

45 

108.0 

145 

S5 

S6.6 

155 

65 

19.0 

165 

75 

175 

85 

\B0 

Fig.  14. — Photometric  distribution  curve  and  data  on  reflector  shown 
in  Fig.  19,  with  a  1.12-watt-per-caudle  lamp. 


wheeler:   the  lighting  of  show  windows 


569 


Fig.  15.— Reflector  (No.  750).    Details  for  installing 
in  window  shown  in  Fig.  16. 


I-Condulef. 
Z-Pubber  Gasket. 
3-  Receptacle. 
Plug 


Section  and  Plan 
showing  installation 
de  tolls  of 
Reflector  No.    750 


I  40  Walt  Clear  Bulb 
J  Lamp 


Fig.  16. —Details  for  installing  in  windows  reflector 
(No.  750)  shown  in  Fig.  15. 


57o 


TRANSACTIONS  I.  E.   S. — PART  II 


Fig.  17. — Test  window  set  up  as  a  Class  B  window. 


Fig.  18. — A  Class  E  window  ;  height  5  ft.,  3  in.,  depth  3  ft.,  equipped  with  60-watt  clear 
tungsten  lanips  in  reflectors  (Fig.  19)  on  15-inch  centers. 


wheeeer:   the  lighting  oe  show  windows  571 


Fig.  19.— Reflector  (No.  696).    Details  for  installing 
in  windows  shown  in  Fig.  20. 


60  Watt  Cleat- Bulb 
Reflector  A/o  696.  Lamp 

(Unskirted  base. ) 
Fig.  20.— Details  for  installing  in  windows  reflector  (No.  696)  shown  in  Fig.  19. 


Fig.  21. — Show  window  drape  (valance). 


572 


TRANSACTIONS  I.  E.   S. — PART  II 


CEILING  LINEv 

u  / 

< 

I4\ 

14 

r 

13  \ 

13 

»2\ 

12 

"\ 

II 

UJ 
UJ 

® 

?.\ 
8 

7 

6 

10 

. -*^- 

_S^. 

X- 

^ 

k..-^S 

(c>N 

^- 

-— ^ 

8 

Q 

g 

UJ 
1 

1 

1/  o 
1  < 

1  u. 
\  o 

■fc 

is 

x 

to 
•z. 

6 

u. 

o 
o 
•z. 

5 
4- 

5 

\    4 

3 

o 

o 

2 

2 

X 

-NO.  780 
-NO.  755 

u? 

1 

1 

UJ 

DC 

^V^" 

i 

2 

3 

4 

5 

6 

7 

8 

s  i< 

UJ      > 

CO     / 

FLOORS               (§) 

! 

Fig.  22. — Chart  for  the  selection  of  the  proper  reflector. 


wheeler:  the  lighting  oe  show  windows  573 

SELECTION  OF  REFLECTORS. 

The  chart  shown  in  Fig.  22  makes  easy  the  selection  of  the 
correct  reflector  for  any  type  of  window.  To  start  with,  knowl- 
edge of  three  things  is  necessary :  height,  depth  of  window,  and 
height  of  trim  at  background. 

Suppose,  for  example,  the  show  window  is  10  ft.  (3.04  m.) 
high,  6  ft.  (1.83  m.)  deep  and  the  trim  or  background  to  be 
lighted  is  6  ft.  (1.83  m.)  high.  The  procedure  would  be  as 
follows : 

First,  find  the  height  of  the  window  (10  ft.)  on  the  left-hand 
vertical  scale.    This  point  is  indicated  at  "A." 

Second,  locate  depth  of  window  (6  ft.)  on  lower  horizontal 
scale.     This  point  is  indicated  at  "B." 

Third,  straight  up  from  the  latter  point  locate  another  point 
corresponding  to  the  highest  point  to  which  the  window  is 
trimmed  (in  this  case  6  ft.).     This  point  is  indicated  at  "C." 

Next  note  the  diagonal  line  that  most  nearly  passes  through 
the  two  points,  "A"  and  "C"  (which  in  this  case  is  a  heavy  dark 
line).  By  referring  to  the  key  below  one  finds  the  reflector 
designated  by  the  heavy  dark  line  — — — ■— — ^ — — ^— 
is  reflector  No.  780,  which  is  the  correct  reflector  for  this  win- 
dow. In  the  above  example,  if  the  window  were  12  ft.  high,  the 
chart  would  call  for  reflector  No.  755.  This  chart  is  based  on 
the  use  of  the  lamp  for  which  each  reflector  herein  mentioned  is 
designed. 

Where  it  is  necessary  to  place  reflectors  on  the  transom  bar, 
they  may  be  selected  by  using  the  distance  from  the  floor  of  the 
window  to  the  bar  as  the  height  of  ceiling. 


SPACING  OF  REFLECTORS.* 

For   Class   A-B-C   windows   the   spacing,   or   distance    from 
center  to  center  for  the  respective  reflectors,  is  about  as  follows : 

*  For  Class  D  and  E  sec  tables  7  and  8  respectively  for  average  conditions. 


574 

TRANSACTIONS  I.  I 

J.  S. — PART 

II 

Population  of 
town  or  city 

TABLE  I.—  Refxector  Spacings. 

Size  of 
Reflector                    Spacing         lamp  watts 

Average 
foot-candles 

along 
line  of  trim 

10,000  and  under 

No.  777  (Fig.  2) 

28  in. 

60 

10 

No.  780  (Fig.  7) 

36  in. 

100 

10 

No.  755  (Fig.  10) 

48  in. 

100 

10 

10,000-30,000 

No.  777  (Fig.  2) 

18  in. 

60 

15 

No.  780  (Fig.  7) 

24  in. 

100 

15 

No.  755  (Fig.  10) 

36  in 

100 

15 

30,000  and  up 

No.  777  (Fig.  2) 

10  in. 

60 

30 

No.  780  (Fig.  7) 

12  in. 

100 

30 

No.  755  (Fig.  10) 

18  in. 

100 

30 

The  above  table  gives  average  illumination  intensities  found 
good  practise  in  the  live  business  centers  of  a  town.  For  stores 
on  the  outskirts  and  in  the  outlying  districts,  reduced  intensities 
will  often  be  ample.  Also  due  allowance  must  be  made  when  the 
interior  finish  and  trim  is  darker  or  lighter  than  the  light  oak 
for  which  the  table  has  been  compiled. 

LAMPS. 
The  window  reflectors  discussed  in  this  paper  are  designed 
exclusively  for  use  with  the  standard  tungsten  lamps.  Each 
reflector  is  designed  for  a  certain  size  of  lamp.  All  test  data 
included  in  this  paper  is  based  on  the  use  of  the  proper  sizes  of 
lamps  for  which  the  reflectors  are  designed. 

REFLECTOR  HOLDERS. 

Owing  to  the  fact  that  the  many  makes  of  shade  holders 
differ  materially  in  dimensions,  I  have  found  it  essential  to  meas- 
ure all  known  makes  and  determine  which  of  these  insure  the 
proper  position  of  lamp  with  respect  to  the  reflector.  It  is  quite 
essential  that  this  position  should  be  correct,  if  the  best  results 
are  desired. 

Fig.  23  shows  a  portion  of  a  lamp,  reflector,  and  holder.  The 
values  of  dimensions  F  and  G  for  the  various  reflectors  are  tabu- 
lated below,  and  the  trade  names  of  the  shade  holders  having 
these  dimensions  are  given. 

If  all  shade  holders  were  standardized  it  would  be  of  infinite 
value  to  all  concerned. 


wheeler:  the  lighting  of  show  windows 


575 


TAB 

LE  II. 

Porcelain 

Brass 

F 

G 

receptacles 

shell  sockets 

2%  ins. 

i  in. 

A-C 

B-H-X-S-P 

2%  ins. 

I  15/16  in. 

A-C 

B-H-P 

1%  ins. 

1  15/16  in. 

A 

H 

2%  ins. 

1  in. 

A-C 

B-H-X-S-P 

Appleton ; 

C--Crouse-Hinds  ;       H— Holophane 

Reflector    Holder  position 

Fig.  2  Form  "O" 

Fig.  7  "      "H" 

Fig.  10  "      "A" 

Fig.  19  "      "O" 

Symbols  used  :       A — 

B— Bryant ;  P— Plume  &  Atwood  ;    X-Hubbel  ;    S— P.  &  S. 


Fig.  23.— Diagram  of  reflector  and  lamp  socket. 

WINDOW  DRAPES  (VALANCES). 
In  order  to  conceal  show  window  reflectors  from  the  view  of 
the  observer  on  the  street,  window  drapes  or  valances  are  very 
much  used.  In  addition  to  performing  the  above  functions,  they 
add  a  touch  of  refinement  and  exclusiveness  to  a  show  window, 
greatly  increasing  the  attractiveness  of  the  merchandise  dis- 
played. Many  attractive  designs  are  now  being  manufactured 
for  this  use.  Fig.  21  shows  one  type  in  which  the  firm's  mono- 
gram is  inscribed. 

TEST  METHODS  AND  DATA. 

Test  Windows. — Experimental  windows  for  use  in  testing  the 
illumination  of  Class  A,  B,  C,  D  and  E  windows,  were  made 
from  wood  frames  of  various  sizes,  over  which  was  stretched  a 
heavy  paper  backed  with  cloth,  in  much  the  fashion  of  stage 
scenery.  This  paper  was  the  color  of  natural  light  oak.  By 
using  various  combinations  of  the  frames  at  hand,  the  typical 
windows  were  readily  constructed.  For  the  floors  the  same  type 
of  paper  in  a  darker  oak  was  used. 

Fig.  17  shows  the  test  window  set  up  as  a  class  B  window, 
with  reflectors  like  that  of  Fig.  7  in  place.  From  the  illustration 
it  will  be  noted  that  the  reflectors  and  lamps  are  supported  from 
the  2  in.  x  4  in.  (5.08  cm.  x  10.16  cm.)  strip,  which  was  supported 


576  TRANSACTIONS  I.  E.  S. — PART  II 

from  two  standards  having  raising  and  lowering  features.  Pull 
sockets  on  6-inch  (15.24  cm.)  centers  the  entire  length  of  the 
wood  strip,  made  it  possible  to  obtain  the  various  reflector  spac- 
ings  desired. 

A  black  curtain  was  hung  at  the  front  of  the  window  in  the 
position  of  the  glass  in  an  actual  window.  In  windows  having 
a  medium  height  of  trim  in  which  the  upper  portion  of  the 
background  is  glass,  as  for  instance  Class  B,  a  black  curtain  was 
hung  in  this  position.  The  black  cloth  which  reflects  a  negligible 
amount  of  light  flux,  was  considered  a  good  substitution  for 
clear  glass.  Whatever  slight  difference  this  substitution  may 
make,  I  feel  it  is  of  small  moment  and  that  the  results  obtained 
are  representative  of  the  various  classes  of  windows  tested. 

The  line  of  trim  for  the  various  classes  of  windows  was  chosen 
for  the  condition  I  found  existing  in  the  majority  of  windows. 

Instruments. — The  instruments  used  for  this  series  of  tests 
were  a  Sharp-Millar  portable  photometer  with  a  standard  lamp 
and  detached  test  plate  of  white  diffusing  glass  mounted  on  black 
velvet;  a  Weston  milli-ammeter  with  shunt  and  shunt  leads;  and 
a  Wagner  A.  C.  or  D.  C.  voltmeter.  All  instruments  were  care- 
fully calibrated,  both  before  and  after  tests.  The  mean  hori- 
zontal candle-power  and  wattage  of  all  lamps  used,  were  deter- 
mined before  beginning  the  tests.  The  milli-ammeter  was  used 
to  measure  the  current  passing  through  the  standard  lamp.  The 
volt  meter  was  used  to  measure  the  voltage  at  the  lamp  terminals. 

Method  of  Conducting  Tests. — Tests  of  the  normal  illumina- 
tion along  the  line  of  trim  were  made  at  the  stations  as  indicated 
for  each  class  of  window  in  the  subsequent  figures.  The  loca-' 
tion  of  the  stations  was  accomplished  by  measuring  the  hori- 
zontal distances  from  front  to  rear  of  window,  and  marking  the 
vertical  distances  on  the  adjustable  standard  used  for  supporting 
the  detached  test  plate. 

The  photometer  was  placed  in  such  a  position  that  the  line  of 
sight  of  the  instrument  would  be  practically  normal  to  the  line  of 
trim  at  each  station.  The  reason  for  this  was  to  facilitate  plac- 
ing the  test  plate  parallel  with  the  line  of  trim  at  any  one  station, 
since  it  was  found  that  with  a  little  practise  the  operator  was  able 
to  locate  by  eye  the  test  plate  normal  to  the  line  of  sight  of  the  in- 


wheeler:  the  lighting  oe  show  windows  577 

strument.  By  experiment  it  was  also  found  that  it  is  not  necessary 
to  view  the  test  plate  normally.  In  fact  readings  when  viewing 
the  test  plate  at  angles  of  45  degrees  either  side  of  the  normal, 
show  but  slight  difference  from  the  normal  reading.  Hence, 
readings  of  normal  illumination  along  the  line  of  trim  taken 
with  the  photometer  at  the  ordinary  height  of  the  eye,  would  not 
materially  differ  from  those  presented. 

Readings  of  the  voltage  at  the  lamp  terminals  were  made 
simultaneously  with  the  illumination  intensity  readings. 

In  order  to  obtain  some  idea  as  to  the  relative  co-efficient  of 
reflection  of  the  walls,  ceiling  and  floor  of  the  test  windows,  the 
following  test  was  conducted : 

The  test  plate  used  throughout  the  test,  was  set  up  in  a  given 
position,  under  the  illumination  from  tungsten  lamps,  and  the 
intensity  of  illumination  incident  thereon  read. 

The  test  plate  was  replaced  by  a  disk  of  white  blotting  paper 
and  another  reading  taken  with  the  photometer.  The  white 
blotting  paper  was  in  turn  replaced  by  a  sample  of  the  paper  used 
on  the  walls  and  floor  of  the  test  window. 

Assuming  a  reflection  co-efficient  of  100  per  cent,  for  the  test 
plate,  the  reflection  co-efficient  of  the  blotting  paper  and  walls 
and  floor  of  the  test  window,  in  terms  of  the  white  test  plate  was 
determined. 

The  values  obtained  are  as  follows : 

1 :  Test  plate  100     per  cent. 

2 :  Blotting  paper  100     per  cent. 

3 :  Walls  of  test  window 58.4  per  cent. 

4 :  Floor  of  test  window 34.6  per  cent. 

Later  the  absolute  co-efficient  of  reflection  of  the  above  blott- 
ing paper1  was  found  to  be  75  per  cent.2  From  this  it  follows 
that  the  absolute  co-efficient  of  reflection  of  the  above  surfaces 
are  approximately  (limited  by  the  accuracy  of  the  test)  as 
follows : 

Per  cent. 

test  plate   75.0 

Blotting  paper   (see  Notes  below)    75.0 

Floor  of  test  window  25.9 

Walls  of  test  window   43.8 

1  140  Star. 

2  Determined  by  Mr.  M.  ^uckiesh,  of  National  Electric  I^anip  Association. 


57§  TRANSACTIONS  I.  E.   S. — PART  II 

These  values  of  course  are  not  the  absolute  reflection  co-effi- 
cients of  the  surfaces,  but  may  be  used  comparatively,  inasmuch 
as  it  is  known  that  white  blotting  paper  reflects  about  75  to  82 
per  cent,  light. 

Method  of  Calculating  Test  Results. — In  calculating  the  test 
results  of  illumination,  foot-candle  readings  were  corrected  for 
115  volts,  the  rated  high  voltage  of  the  lamps  used.  These  cor- 
rections were  made  in  accordance  with  the  lamp  data  supplied 
by  the  lamp  manufacturer. 

The  effective  lumens  on  the  surface  of  the  line  of  trim  were 
calculated  from  the  average  of  all  the  foot-candle  readings  taken 
in  any  one  window.  The  actual  wattage  and  mean  horizontal 
candle-power  of  all  lamps  being  known  for  115  volts,  the  effec- 
tive lumens  per  watt  and  the  efficiency  of  utilization  were  calcu- 
lated. This  data  is  presented  in  the  illumination  tables  for  each 
class  of  window  tested. 

A  floor  plan  and  an  elevation  at  one  of  the  planes  tested,  is 
shown  with  the  data  on  each  of  the  windows.  On  the  floor  plan 
are  indicated  the  location  of  the  test  planes  and  the  location,  num- 
ber and  spacing  of  the  outlets.  On  the  elevation  are  shown  the 
line  of  trim,  the  test  stations,  and  the  reflector  showing  its  po- 
sition relative  to  the  line  of  trim. 

In  addition  to  this  a  curve  showing  the  normal  illumination 
at  the  various  test  stations  is  given.  The  method  of  plotting 
this  curve  is  rather  peculiar,  inasmuch  as  it  is  neither  a  polar  nor 
a  rectangular  co-ordinate  curve.  The  line  of  trim  is  indicated 
as  the  zero  line  of  illumination  intensity,  and  the  various  lines 
parallel  to  the  line  of  trim  are  given  foot-candle  values  as  indi- 
cated. The  length  of  the  ordinate  from  the  line  of  trim  to  the 
curve,  represents  the  foot  candle  intensity  of  normal  illumination 
at  the  station. 

This  curve  drawn  as  it  is,  I  believe  is  contrary  to  all  precedent, 
and  while  it  cannot  be -used  for  obtaining  close  approximations 
of  the  illumination  intensity  at  points  between  stations,  I  believe 
it  is  a  good  method  for  giving  one  a  graphical  idea  of  the  distri- 
bution of  illumination  intensity  along  the  line  of  trim. 


wheeler:   the  lighting  of  show  windows 


579 


Fig.  24. — Elevation  through  Class  A  window 
showing  uniformity  of  normal  illumina- 
tion along  the  line  of  trim. 


Fig.  25.— Plan   of  Class  A  window   showing 
location  of  test  planes  and  ceiling  outlets. 


TABLE    3.— Showing  Test  Re- 
sults  in    Class    A   Window 
with     Reflector     No.     777 
Shown  in  Fig.  2. 
Centers,  18  in. 
Lamps,   60- watt,    tungsten  clear 

skirted. 

Total  watts,  501.9. 

Number  of  units,  8. 

Holder,  2%  in.  "O." 

Watts  per  running  foot,  41.82. 

Average 

foot- 
candles 


Plane 
A.... 


Station 
.  ..    I 
2 

3 
4 
5 
6 

7 
8 

9 
.  •   1 

2 
3 


12.4 
16.0 
19.2 
16.3 

17.7 
18.5 
23-7 
22.7 

17.4 
10.0 

13-9 
16.6 


Plane 
B.... 


Station 
...    4 

5 
6 

7 
8 

9 

. . .    1 


Average 

foot- 
candles 

12.3 

14.9 

17.I 

19.8 

19-3 
14.2 

13-7 
16.3 

15.9 
13-4 

11. 6 
12.2 
14.0 
13.S 
1 1 -5 


Average  foot-candles,  15.75 

Area     of     surface    along    line    of 

trim.  168  sq.  ft. 
Effective  lumens,  2650. 
Effective  lumens  per  watt,  5.27. 
Total  lumens  of  lamps  alone,  4533. 
Efficiency  of  utilization,    58.3  per 

cent. 


58o 


TRANSACTIONS  I.  E.   S. PART  II 


Fig.  26. — Elevation  through  Class  B  window 
(medium  trim)  showing  uniformity  of 
normal  illumination  along  line  of  trim. 

TABLE  4.— Showing  Test  Re- 
sults in  Class  B  Window 
(  Medium  Trim  )  with  Reflec- 
tor No.  780,  Shown  in  Fig.  7. 

Centers,  24  in. 

Lamps,  100-watt  clear  tungsten. 

Total  watts,  501.5. 

Number  of  units,  5. 

Holder,  3^  in.  "A." 

Watts  per  running  foot,  50.15. 


Plane 

Station 

Average 

foot- 
candies 

A 

24.4 

2 

21.7 

3 

23.I 

4 

2I-3 

5 

19-5 

6 

.10.0 

B 

7 

2.3 

20.1 
21.7 

2 

3 

24.2 

PIRN 


1_ 


Fig.  27. — Plan  of  Class  B  window  (medium 
trim,)  showing  location  of  test  planes 
and  ceiling  outlets. 


Plane 

Station 

Axerage 

foot- 
candles 

B 

4 

18.8 

5 

22.3 

6 

II. I 

7 

2.8 

c 

17.8 
17.7 

2 

3 

17.0 

4 

i5-i 

5 

15-2 

6 

11. 8 

7 

3-o 

Average  foot-candles,  16.22. 

Area    of    surface     along    line    of 

trim,  95.0  sq.  ft. 
Effective  lumens,  1540. 
Effective  lumens  per  watt,  3.08. 
Total  lumens  of  lamps  alone,  4745. 
Efficiency  of  utilization,  32.5  per 

cent. 


y*x 


\ei~-/7£  "A-  /4i  "■+>  aj  '^>  10  i 


~« 

6 

O 

' 

-0" 

S" 

>- 

PLAIN.                  % 

r 

sNO.  780  REFLECTOR   % 
&  WITH  I0OW.  LAMP    £ 

-^-    &N0.  777  REFLECTOR    & 
WITH  60  W.  LAMP       § 

s 

4 

PLANS  A' 

O 

<V 

+ 

PLANE  ft 

i 

4 

♦  1 

QO    PLANE  8 

t   ) 

PLANE  C 

5 

,T-?- 

4i 

T 

1    I 

• 

Fig.  28. — Elevatioti  through  Class  B  window 
(high  trim)  showing  uniformity  of  nor- 
mal illumination  along  the  line  of  trim. 

TABLE  5.— Showing  Test  Re- 
sults in  Class  B  Window 
(High  Trim)  with  a  Com- 
bination of  Reflectors  Nos. 
780  and  777  Shown  in  Figs.  7 
and  2  Respectively. 
Centers,  18  in. 
Lamps,  4  100-watt  clear  tungsten 

skirted. 

Lamps,  4  60-watt  clear  tungsten 

skirted. 

Total  watts,  653.55. 
Number  of  units,  4  No.  7S0. 
Number  of  units,  4  No.  777. 
Holder,  No.  780,  3^  in.   "A." 
Holder,  No.  777,  2%  in.  "  O." 
Watts  per  running  foot,  54.46. 

Average 
Plane                         Station      foot-candles 
A 1  21.9 


Plane 
Al    . 


B 


A 1 


24.2 
22.6 
27.7 
22.5 
18.3 
18.3 
17.9 
20.5 
28.1 
24.8 
21.0 


Fig.  29. — Plan  of  Class  B  window  (high 
trim)  showing  location  of  test  planes 
and  ceiling  outlets. 

Average 

foot- 
candles 
20.3 
15-6 
16.5 
23.2 
20.3 
22.9 
23.2 
21.0 
18.9 
16.3 
13-6 
14.8 
19.10 
21.3 
22.1 
21.8 
21.4 
16.2 

151 
8  10.4 

Average  foot-candles,  20.23. 

Area   of  surface    along    line    of 
trim,  120.00. 

Effective  lumens,  2430. 

Effective  lumens  per  watt,  3.73. 

Total   lumens    of    lamps    alone, 
6023. 

Efficiency  of  utilization,  40.3  per 
cent. 


Station 
••    5 
6 

7 
8 
..  1 
2 
3 
4 
5 
6 

7 
8 
. .  1 
2 
3 
4 
5 
6 

7 


3 


582 


TRANSACTIONS  I.  E.   S. — PART  II 


[6i~/6£^k-/3"X. 


Fig.  30. — Elevation  of  Class  C  window  show- 
ing uniformity  of  normal  illumination 
along  the  line  of  trim. 

TABLE   6.— Showing  Test  Re- 
sults  in    Class    C    Window 
with     Reflector     No.     755 
Shown  in  Fig.  10. 
Centers,  36  in. 

Lamps,  100-watt  tungsten  clear. 
Total  watts, ^409.6.  J 
Number  of  units,  4. 
Holder,  $%  in.  "O." 
Watts  per  running  foot,  34.13. 


Plane 

Station 

Average 

foot- 
candles 

21.8 

2 

22.2 

3 

18.4 

4 

14.6 

5 

12.6 

6 

7-7 

7 

4.1 

B 

19.7 

2 

22.9 

PL /J  A/ 


<K 


PLANE  A. 


PLANE  8. 


1 t      i      I 


PLANE  C 


Fig.  31. — Plan  of  Class  C  window  showing 
location  of  test  planes  and  ceiling 
outlets. 


Average 

foot- 

Plane 

Station 

candles 

B 

21.8 

4 

l6.2 

5 

u-3 

6 

7-3 

7 

4.1 

c 

1 

16.5 

2 

18.4 

3 

18.0 

4 

15-4 

5 

12.3 

6 

7-9 

7 

4-3 

Average  foot-candles,  14.16. 

Area  of  surface  along  line  of 
trim,  114  sq.  ft. 

Effective  lumens,  1610. 

Effective  lumens  per  watt,  3.93. 

Total  lumens  of  lamps  alone, 
3862. 

Efficiency  of  utilization,  41.7. 


wheeler:   the  lighting  of  show  windows 


583 


!> 


NO.750 REFLECTOR  JOOT^ 


l>6t*-  /8"4-  174  "X*  i4i^»  n'Xsk 


Fig.  32. — Elevation  of  Class  D  window  show- 
ing uniformity  of  normal  illumination 
along  the  line  of  trim. 


Fig-  33— Plan  of  Class  D  window  show- 
ing location  of  test  planes  and  ceiling 
outlets. 


TABLE  7.— Showing  Test  Re- 
sults in  Class  D  Window 
with  Reflector  No.  750 
Shown  in  Fig.  15. 

Centers,  24  in. 

Lamps,    60-watt  tungsten    clear 
short  base. 

Total  watts,  623.4. 
Number  of  units,  5. 
Holder,  special. 
Watts  per  running  foot  62.34. 


Plane 

Station 

Average 

foot- 
candles 

A 

34-2 

2 

37-6 

3 

45-o 

4 

42.8 

5 

37-2 

6 

26.0 

B 

38-6 

2 

43-6 

Plane 

Station 

Average 

foot- 
candles 

B 

3 

4 

47-5 
46.7 

c 

5 
6 

35-5 
25.8 
26.9 
38.6 

2 

3 

43- 0 

4 

37-3 

5 
6 

33-4 
22.7 

Average  foot-candles,  36.8. 

Area  of  surface  along  line  of 
trim, "86. 24  sq.  ft. 

Effective  lumens,  3170. 

Effective  lumens  per  watt,  5.08. 

Total  lumens  of  lamps  alone, 
5642. 

Efficiency  of  utilization,  56.2  per 
cent. 


584 


TRANSACTIONS  I.  E.   S. — PART  II 


Fig.  34.— Elevation  of  Class 
E  window  showing  uni- 
formity of  normal  illum- 
ination along  the  line  of 
trim. 

TABLE  8.— Showing  Test  Re- 
sults   in    Ci,ass    E   Window 
with    no.    696    reflectors, 
Shown  in  Fig.  19. 
Centers,  24  in. 

Lamps,   60-watt  tungsten    clear 
short  base. 
Total  watts,  372.45. 
Number  of  units,  6. 
Holders,  2%  in.  "  O." 
Watts  per  running  foot,  31.04. 


Plane 

Station 

Average 

foot- 
candles 

A 

30.0 

2 

32.9 

3 

28.3 

4 

28.O 

5 

20.3 

1 

25.5 

[— 

Z:6 

■^          * 

i 

| 

PLPN.  § 

Si 

As                 G 

$~ 

PL  ONE  fl 

O 

+ 

PLPNE  B. 

V 

CO 

t" 

"*■ 

» 

<0 

* 

M- 

J 

PLANE  C 

<\\ 

5 

\ 

<P 

^ 

| 

t 

\ 

" 

Fig.  35. — Plan  of  Class 
E  window  showing 
location  of  test 
planes  aud  ceiling 
outlets. 


Plane 

Station 

foot- 
candles 

B 

30.I 

3 

26.2 

4 

30-5 

5 

19-3 

c 

18.6 
23.6 

2 

3 

21.7 

4 

21.8 

5 

16.2 

Average  foot-candles,  24.8. 
Area    of    surface  along   line  of 
trim,  40  sq.  ft. 

Effective  lumens,  992. 
Effective  lumens  per  watt,  2.67. 
Total  lumens   of    lamps    alone, 

3349- 

Efficiency  of  utilization,  29.7  per 
cent. 


wheeler:   the  lighting  of  show  windows  585 

The  results  obtained  in  Class  B  and  C  windows  I  believe  are 
best.  A  high  intensity  of  illumination  at  the  bottom  and  front 
of  the  window  with  a  gradual  decrease  up  to  the  top  of  trim 
seems  to  give  the  desired  stage  light  effect  in  this  class  of  win- 
dows. 

In  the  Class  A,  D  and  E  windows  the  illumination  intensity 
along  the  line  of  trim  is  nearly  constant.  A  very  good  effect  is 
secured  in  each  case. 

Tests  in  the  various  windows  run  with  the  reflectors  with  var- 
ious spacings,  as  would  be  expected,  showed  that  the  average 
intensity  of  illumination  along  the  line  of  trim  varies  practically 
inversely  with  the  spacing  of  the  units. 

The  tests  in  Class  A,  B  and  C  windows  reported  here,  are 
representative  of  the  intensities  I  have  recommended  for  the 
average  size  city.  Windows  of  the  classes  D  and  E  are  more 
prevalent  in  larger  cities  and  hence  the  tests  reported  here  show 
higher  intensities  of  illumination. 

The  chart  for  determining  the  type  of  reflector  required  for 
a  given  window,  and  the  table  giving  the  spacings  for  the  various 
reflectors,  check  out  with  the  test  results. 

The  reason  that  the  No.  JJJ  and  No.  750  (Figs.  2  and  15 
respectively)  reflectors  show  efficiencies  of  utilization  consider- 
ably above  any  of  the  other  reflectors,  is  evident  because  the 
area  in  which  the  light  is  to  be  distributed  is  large  and  takes 
in  a  wide  angle.  No  difficulty  is  experienced  in  confining  the 
maximum  flux  from  the  lamp  in  this  area. 

In  the  case  of  the  No.  755,  No.  780  and  No.  696  reflectors, 
(Figs.  10,  7  and  19  respectively),  the  area  in  which  the  light 
flux  is  to  be  directed  is  relatively  smaller,  and  takes  in  a  much 
smaller  angle.  It  is  this  fact  that  causes  the  resultant  lower 
efficiencies  of  utilization.  Of  course  it  would  not  be  desirable 
to  cut  off  all  light  outside  the  angle  subtended  by  the  line  of 
trim,  but  for  the  best  effect  and  most  efficient  results  the  largest 
portion  of  the  light  flux  must  be  confined  to  this  angle. 

The  No.  696  reflector  (Fig.  19)  as  has  been  previously  ex- 
plained, allows  too  much  light  to  escape  outside  of  the  angle  sub- 
tended by  the  line  of  trim.    A  class  E  window  equipped  with  a 


586  TRANSACTIONS  I.  E.  S. — PART  II 

reflector  having  a  distribution  of  light  like  that  of  Fig.  10,  would 
show  a  much  better  efficiency  of  utilization. 

The  reflector  shown  in  Fig.  7  should  not  show  a  lower  efficiency 
of  utilization  than  the  reflector  of  Fig.  10.  The  difference  shown 
by  the  test  is  due  to  the  fact,  that  the  No.  780  (Fig.  7)  reflectors 
tested,  happened  to  be  the  first  ones  taken  from  the  mold. 
This  mold  after  a  few  reflectors  have  been  run,  becomes  much 
smoother  and  the  later  reflectors  show  a  very  much  increased 
efficiency.  The  efficiencies  of  the  latter  two  reflectors  used  for 
these  tests  have  been  determined  and  show  approximately  the 
same  difference. 

In  order  to  obtain  an  idea  as  to  how  much  the  walls  and  ceiling 

of  the  windows  add  to  the  total  illumination  on  the  line  of  trim 

by  reflection,  tests  were  run  on  a  class  A  window  with  the  walls, 

ceiling  and  floor  covered  with  black  cloth.    The  conditions  of  the 

previous  test  in  the  class  A  window  were  duplicated  with  the 

exceptions   noted   above.     The   result  of   these   tests   were   as 

follows : 

Average   foot-candles    1470 

Area  of  surface  along  line  of  trim 168  sq.  ft. 

Effective  lumens   2470 

Effective  lumens  per  watt 4.92 

Total  lumens  of  lamps  alone 4546 

Efficiency  of  utilization  54.3  per  cent. 

Comparing  these  figures  with  those  on  the  light  oak  window, 
it  is  apparent  that  7.37  per  cent,  light  flux  is  reflected  onto  the 
line  of  trim  by  the  walls  and  ceilings  of  the  window.  This  added 
illumination  will  be  more  or  less  depending  on  the  type  of  window 
and  interior  finish.  The  light  oak  window  was  chosen  as  a  good 
average  finish. 

CONCLUSION. 
As  noted  at  the  beginning  of  my  paper,  the  subject  of  win- 
dow lighting  has  received  only  a  limited  amount  of  attention 
from  the  society,  and  I  trust  I  have  started  something  which  will 
lead  to  more  investigations  in  this  direction.  It  is  a  subject  well 
worthy  of  the  attention  of  the  illuminating  engineer,  the  re- 
flector manufacturer,  the  lamp  manufacturer,  the  electrical  con- 
tractor, and  the  central  station.    It  is  always  possible  to  interest 


wheeler:   the  lighting  of  show  windows  587 

a  merchant  in  the  true  advertising  value  of  a  well  illuminated 
window  display,  because  he  can  see  increased  business.  It  is  a 
monetary  consideration  with  him.  He  is  open  to  conviction 
when  the  question  of  increasing  sales  is  under  consideration. 

In  addition  to  this,  however,  beautifully  illuminated  show  win- 
dows enhance  the  beauty  of  a  city,  and  thus  will  receive  the 
endorsement  of  the  populace  as  a  public  benefit. 

I  wish  to  acknowledge  valuable  suggestions  in  preparing  this 
paper  from  Messrs.  J.  R.  Cravath,  J.  B.  Jackson,  and  L.  V. 
James,  associate,  electrical  engineering  department,  University 
of  Illinois,  and  for  the  assistance  of  the  engineering  department 
of  the  National   X-Ray  Reflector   Co. 

Note  :  The  reflectors  referred  to  by  number  in  this  paper  are  manufactured  by  the 
National  X-Ray  Reflector  Company,  Chicago,  111. 

DISCUSSION. 

Mr.  J.  R.  Cravath,  Chicago :  I  believe  with  Mr.  Little  that 
the  surface  brightness  in  the  line  of  trim  as  seen  from  the  street 
should  be  the  criterion  by  which  we  should  make  our  installation, 
but  I  also  think  that  Mr.  Wheeler's  method  practically  gives  us 
that  within  a  very  narrow  margin  of  error,  because  we  can 
see  by  his  method  that  he  moved  a  diffusing  test  plate  up  and 
down  almost  along  the  line  of  trim  or  very  close  to  it. 

The  main  point  I  want  to  make  is  something  in  regard  to 
architecture  of  show  windows.  It  has  been  said  that  we  must 
take  show  windows  as  we  find  them ;  but  there  are  frequently 
cases  where  a  merchant  is  remodelling  his  show  windows  to 
make  them  more  effective  and  the  illuminating  engineer  ought 
to  be  in  a  position  to  recommend  the  dimensions  which  will 
make  it  possible  to  bring  out  the  goods  to  the  best  advantage 
We  all  know  that  show  windows  are  at  their  best  at  night  under 
artificial  illumination  rather  than  in  the  day  time;  therefore 
their  proportions  should  be  selected  or  should  be  designed  to 
give  the  best  results  under  artificial  -illumination.  Experience 
has  proven  that  the  type  of  window  which  Mr.  Wheeler  classi- 
fies as  a  Class  C,  that  is,  one  with  the  height  two  times  the  depth, 
does  not  give  good  results  with  artificial  light ;  the  result  simply  be- 
ing that  the  vertical  illumination  that  shines  on  the  goods  is  com- 
paratively low.    This  applies  to  dry  goods.    Of  course  with  jew- 


588  TRANSACTIONS  I.  E.  S. — PART  II 

elry  it  is  altogether  different.  It  is  almost  impossible  to  avoid 
having  dry  goods  cast  shadows  on  themselves  in  such  shallow 
windows.  I  prefer  for  dry  goods  a  deep  window  as  that  type 
makes  possible  the  best  illumination  of  the  goods. 

Mr.  W.  F.  Little)  :  In  addition  to  this  very  valuable  paper, 
would  it  not  be  well  to  make  further  investigations  showing 
the  surface  brightness  of  the  various  objects  in  the  line  of  trim? 
While  the  foot-candle  intensities  as  given  by  Mr.  Wheeler  are 
of  prime  importance,  still  surface  brightness  measurements 
showing  the  range  of  contrasts  met  with  in  a  window  trimmed 
with  various  materials  would  be  of  considerable  interest. 

Mr.  H.  B.  WhEEEER  (In  reply)  :  Regardless  of  the  class  a 
window  falls  into,  the  practise  of  using  mirrored  backgrounds 
should  be  discouraged,  as  images  of  surrounding  articles  are  gen- 
erally present,  which  detract  from  the  goods  on  display. 

Under  typical  windows  in  the  large  cities  on  the  main  streets, 
the  merchant  is  very  anxious  to  obtain  the  full  advertising  value 
of  his  window ;  and  hence  demands  a  high  degree  of  illumination. 
Thirty-foot  candles  has  been  found  ample  illumination  for  typical 
windows  with  medium  decorations. 

From  the  data  presented  in  this  paper,  you  will  note  that  con- 
centrating reflectors  are  used  in  comparatively  high  windows, 
and  hence  are  at  a  considerable  distance  from  the  goods  on  dis- 
play. Further  the  reflectors  are  designed  to  allow  plenty  of 
radiation.  I  have  never  heard  of  any  case  where  the  light  rays 
from  commercial  types  of  concentrating  reflectors,  had  ever  done 
any  damage  in  the  least  to  the  most  delicate  goods  in  a  window. 

A  further  investigation  of  surface  brightness  would  be  very 
interesting,  and  I  trust  that  further  investigation  on  this  subject 
will  be  gone  into. 

My  paper  was  based  on  average  conditions  typical  of  fairly 
modern  show  windows.  Medium  colors  were  selected  as  fairly 
representative  for  the  ends,  background  and  ceiling  of  the 
various  windows  tested. 


HIBBEN  :     STREET    RAILWAY    ILLUMINATION  589 

MODERN  PRACTISE  IN  STREET  RAILWAY 
ILLUMINATION.* 


BY    S.    G.    HIBBEN. 


Synopsis:  Up  to  the  present  time  street  car  lighting  has  been  done 
inefficiently,  with  bare  carbon  filament  lamps.  Recently  four  standard 
tungsten  lamps  have  been  placed  on  the  market,  which  in  conjunction 
with  proper  downward  reflecting  shades  have  enabled  the  energy  cost  for 
lighting  to  be  cut  in  half,  at  the  same  time  allowing  an  increase  in  the 
usable  light  of  more  than  80  per  cent.  Special  fixtures  are  available  for 
supporting  the  shades,  and  selector  switches  may  be  employed  to  insure 
continuous  lighting  service  of  series  burning  lamps.  Steadier  illumina- 
tion is  secured  by  the  new  system  of  lighting,  under  the  adverse  condi- 
tions of  voltage  fluctuations,  and  the  glare  that  exists  at  present  in  the 
majority  of  electrically  lighted  street  cars  is  done  away  with  through 
the  medium  of  the  shades.  The  scientific,  economical  lighting  of  street 
cars  is  a  field  that  is  full  of  interesting  possibilities,  and  is  now  being 
very  rapidly  developed. 

Although  the  incandescent  filament  lamp  has  been  used  for 
the  lighting  of  street  railway  cars  practically  ever  since  the  elec- 
tric motor  cars  superseded  the  horse  or  cable  cars,  it  has  not 
been  until  within  the  last  six  years  that  any  attempts  have  been 
made  to  utilize  the  generated  light  to  best  advantage  by  means 
of  scientifically  manufactured  shades  or  reflectors ;  and  it  has  not 
been  until  about  one  year  ago  that  the  tungsten-filament  lamps  of 
an  efficiency  of  1.4  watts  per  horizontal  candle  or  better,  have 
been  perfected  to  the  extent  of  being  sufficiently  rugged  for 
this  street  railway  service.  Consequently  the  economic  and  scien- 
tific illumination  of  street  cars  has  been  slow  in  its  inception  and 
development,  compared  with  the  rapid  progress  that  has  been 
made  in  the  lighting  of  large  office  buildings,  stores,  residences, 
or  even  steam  railway  cars. 

There  are  several  reasons  for  this  slow  development.  Lighting 
energy,  being  but  a  small  fraction  of  the  total  energy  used  by 
motors,  and  being  relatively  cheap  to  generate,  has  not  been 
considered  as  a  fit  subject  for  economy.    Also  the  short  periods 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society.  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


590  TRANSACTIONS  I.  E.  S. — PART  II 

that  individuals  of  the  traveling  public  use  the  lighting,  and  the 
disposition  of  passengers  to  make  the  best  of  the  illumination, 
unless  it  be  absolutely  unbearable,  has  not  been  conducive  to 
public  protestations.  Furthermore  the  rough  usage  to  which 
lamps  and  shades  are  necessarily  subjected,  the  low  first  cost 
of  the  carbon  filament  lamps,  the  uncertain  relations  between 
private  street  railway  corporations  and  municipalities,  have  all 
been  reasons  for  slowness  in  arriving  at  modern  car  lighting. 
How  really  vital  these  reasons  are  will  be  left  to  be  judged  after 
a  further  consideration  of  this  paper. 

As  far  as  the  author  can  determine,  the  first  street  car  using 
individual  reflectors  on  lamps  was  put  in  service  in  1909,  this 
being  a  car  of  the  Oakwood  Traction  Co.,  operating  in  Dayton, 
Ohio,  although  previously  there  may  have  been  a  few  desultory 
attempts  to  equip  lamp  clusters  with  reflecting  glassware.  This 
car  was  equipped  with  center-deck  4-light  fixtures,  and  side  wall 
single  light  brackets,  using  square  shaped  alba  glass  shades.  A 
view  of  this  car  is  shown  in  Fig.  1.  Cars  with  this  equipment 
are  still  in  service. 

About  two  years  ago  a  number  of  traction  companies  installed 
bare  23  and  36-watt  tungsten  filament  lamps  replacing  the 
carbon  lamps,  having  their  attention  directed  chiefly  towards  de- 
termining the  ultimate  lamp  life.  When  the  majority  of  these 
new  lamps  had  shown  a  life  of  1,000  to  1,300  hours,  the  progress 
was  rapid  towards  the  standardization  of  the  present  series 
burning  lamps,  and  the  shade,  holder,  and  switch  devices  as 
accessories.  The  description  of  such  modern  equipment  and  a 
discussion  of  its  operation  constitutes  the  subject-matter  of  this 
paper. 

PAST  PRACTISE. 

The  lighting  of  street  cars  has  previously  been  accomplished 
by  using  bare  carbon,  and  in  a  few  cars,  graphitized  filament 
lamps.  It  has  required  from  a  dozen  to  thirty  of  the  so-called 
16  candle-power  64-watt  carbon  lamps  in  the  car  body,  and  eight 
to  ten  similar  lamps  distributed  on  platforms  and  in  the  head- 
light and  designating  signs.  Between  bulkheads  the  lamps 
were  placed  about  18  inches  (45.72  cm.)  apart  in  line  along  the 
center  deck,  or  studded  over  the  whole  ceiling,  or  else  grouped 


HIBBEN  :     STREET    RAILWAY    ILLUMINATION  591 

in  clusters  of  four,  five  or  as  many  as  eight  lamps  arranged 
radially  from  single  fixtures  on  the  center  deck  ceiling.  Such 
carbon  lamps  usually  burned  five  in  series  on  the  nominal  550 
volt  power  circuit,  each  being  rated  at  no  volts. 

The  high  current  consumption  of  the  carbon  lamps,  together 
with  their  poor  illuminating  performance,  has  led  to  the  sub- 
stitution of,  first,  the  metallized  or  gem  lamps,  and  second,  the 
bare  23-watt  tungsten  lamps  in  the  same  sockets.  The  former 
lamps  are  proving  unsatisfactory  on  account  of  filament  break- 
age from  jarring  and  other  objections,1  while  the  small  un- 
shielded tungsten  lamps  are  but  a  temporary  makeshift,  due  to 
excessive  glare,  and  because  no  attempt  is  made  to  utilize  the 
maximum  amount  of  generated  light,  or  direct  it  downward. 

PRESENT  METHODS. 

The  most  modern  car  lighting  equipment  consists  of  one 
circuit  of  five  tungsten  lamps,  of  the  94-watt,  78  candle-power 
size,  arranged  in  line  along  the  car  ceiling,  or  else  an  arrange- 
ment of  two  circuits  of  five  each,  of  the  56-watt,  46.7  candle- 
power  tungsten  lamps.  Quite  often,  in  the  cars  where  the  94- 
watt  lamps  are  used,  these  are  placed  four  in  the  car  body 
between  bulkheads,  and  one  over  the  entrance  vestibule,  espec- 
ially if  the  car  is  of  the  pay-enter  type.  In  other  types  of  cars, 
such  as  the  interurbans,  there  may  be  three  units  in  the  passen- 
ger compartment,  one  in  the  baggage  or  smoking  room,  and  one 
in  the  vestibule.  An  additional  circuit  of  five  23-watt  tungsten 
lamps  is  recommended  for  the  large  types  of  city  cars,  particularly 
if  these  cars  have  the  one  circuit  of  94-watt  lamps.  The 
small  lamps  are  arranged  over  the  steps,  in  the  headlight,  and 
in  the  illuminated  designation  signs. 

Sometimes,  but  not  often,  the  fourth  size  of  modern  lamp,  a 
36-watt,  26.8  candle-power  tungsten-filament  is  used  in  the  car 
body,  but  the  cases  where  the  23-watt  or  the  36-watt  lamps  are 
being  employed  between  bulkheads  are  largely  those  where  no 
new  wiring  or  accessories  are  being  installed,  and  where  these 
small  lamps  are  replacing  the  carbon  lamps  in  the  old  sockets 
or  receptacles. 

1  The  average  life  of  Gein  lamps  in  street  railway  service  is  found  to  be  900  hours  or 
less,  although  with  the  exception  of  a  few  isolated  cases,  the  life  is  nearer  700  hours. 


592  TRANSACTIONS  I.  E-   S. — PART  II 

The  four  lamp  sizes  mentioned  above  are  the  ones  that  are  so 
far  standardized  for  street  railway  service.  Their  characteristics 
may  be  found  in  the  following  table : 

TABLE  I.— Characteristics  of  Tungsten-Filament  Street 


Railway  Lamps. 

Watts 

Hor. 
C-p. 

Watts  per 
C-p. 

Avg.  Hrs 
Lumeus             Life 

Bulb 
Diam. 

Overall 
Length 

23 

17.I 

1-34 

168               2,000 

2-78 

5-# 

36 

26.8 

1-34 

263              2,ooo 

2-78 

5-tf 

56 

46.7 

I.20 

457              2,000 

2-5/8 

5-H 

94 

78.3 

I.20 

767               2,000 

3-Vl6 

7-H 

Any  of  these  lamps  are  procurable  for  a  power  line  voltage 
of  525  to  650,  or  with  individual  ratings  of  105  to  130  volts. 
They  are  sturdy  in  construction,  and  selected  for  the  current 
to  insure  a  uniformity  of  candle-power  and  life. 

The  wiring  circuits  of  two  typical  systems  are  given  in  Figs. 
2  and  3.  In  Fig.  2  may  be  seen  the  arrangement  of  the  94-watt 
lamps  that  is  excellent  because  it  insures  continuous  lighting 
service.  An  extra  unit  is  used,  commonly  short-circuited  by  a 
selector  switch  as  shown,  and  thus  in  the  event  of  the  failure  of 
any  one  lamp,  (which  would  leave  the  whole  car  in  darkness), 
the  manually  operated  selector  switch  may  be  turned  by  the 
conductor  or  motorman,  bridging  successive  lamps  until  the 
burnt-out  one  is  short-circuited,  while  meantime  the  extra  unit 
(usually  in  the  rear  vestibule)  comes  into  circuit,  and  the  failure 
may  be  replaced  at  the  first  convenient  point  on  the  run. 
On  the  whole,  this  wiring  arrangement  seems  to  be  the  most 
advisable  one.  This  scheme  is  all  the  more  advantageous  when 
a  three-way  switch  is  wired  in  the  circuit,  and  when  a  seventh 
unit  is  installed  in  the  front  vestibule.  A  car  with  this  arrange- 
ment may  always  have  well  illuminated  steps  and  entrances,  for 
in  loading  or  unloading,  the  conductor  may  bridge  a  unit  in  the 
passenger  compartment,  and  burn  both  vestibule  lights.  Then, 
whichever  way  the  car  is  running,  he  may  so  manipulate  the 
two  switches  as  to  have  a  rear  platform  light,  as  well  as  all 
lamps  between  bulk  heads  burning. 

Fig.  3  shows  an  arrangement  of  two  independent  circuits 
of  56-watt  lamps,  and  since  no  selector  switch  is  used,  any  one 


T 


5W 


Fig-.  4.— One  of  the  modern  units  used  for  street  railway  car  lighting. 


Fig.  5. — View  of  street  car  using  94-watt  tungsten  lamps  and  shades. 


HIBBEN  :     STREET   RAILWAY    ILLUMINATION 


593 


lamp  failure  will  leave  half  but  not  all  of  the  car  in  darkness 
temporarily.  Fig.  ia  shows  a  portion  of  the  rear  platform  with 
a  selector  switch  and  one  ceiling  unit  in  place. 

All  lamps  in  the  most  modern  street  cars  are  being  equipped 
with  downward  reflecting  shades,  for  reasons  hereafter  ex- 
plained. Fig.  4  illustrates  one  such  type  of  shade,  together  with 
the  holder  that  in  reality  is  the  complete  fixture.  Several  forms 
of  holders  are  available ;  one  style  is  shorter  than  the  one  shown 


Fig.  2.— Wiring  diagram  for  a  circuit  of  94-watt  or  56-watt  tungsten  lamps. 


\ 


Sign 

r-O- 


Platform 

Q 

y\       Headlight 


Fig.  3.— Wiring  diagram  for  lamps  in  two  independent  circuits. 


in  Fig.  4.  These  holders  all  operate  to  clamp  the  neck  of  the 
glass  shade  all  around,  with  a  firm  grip  that  cannot  jar  loose, 
and  in  such  a  way  that  there  is  no  probability  of  breakage  if 
a  well  made  shade  is  used. 

Fig.  5  shows  the  interior  of  a  typical  city  car,  where  the  units 
are  of  the  94-watt  type,  with  a  deep  bowl  alba  glass  shade.    The 


594 


TRANSACTIONS  I.  E.   S. — PART  II 


plan  of  this  car  is  given  in  Fig.  6,  from  which  the  location  and 
spacing  of  the  main  lighting  units  can  be  seen. 

ILLUMINATION  COMPARISONS. 

Two  main  factors  are  the  criteria  of  the  satisfactory  qualities 
of  the  lighting  system — the  cost,  and  the  illuminating  perform- 
ance. The  latter  consideration  involves  the  measurable  amount 
of  foot-candles  at  the  desirable  places,  and  also  the  quality  of 
the  light  that  is  furnishing  these  foot-candle  values,  and  its 
physiological  effects. 

Anyone  who  has  seen  a  car  illuminated  by  the  shaded  lamps, 
and  particularly  if  this  car  has  both  shades  and  bare  lamps  that 
can  be  alternately  burned,  will  not  question  the  fact  that  there 


J—36  in*-«33  in^-<33  in~3  3  _in>?«33  inj-^33  in.*7-<33  iiu»f<33in.>p33".in.»f<33  ia-*<33  in-"t<33  in^io  ^fr 
Stations  i  i  i  3  4  5  6  t         e  9  10         II       12     '3  stations 

IBf      Service  Lamp,  90-100  Watt  tungsten 
jsj     Auxiliary  Lamp,  ordinarily  not  bumine 

Fig.  6. — Plan  of  car  shown  in  Fig.  5,  showing  location  of  lamps  and  test  stations. 

is  a  remarkable  difference  in  the  qualities  of  the  light  from  the 
two  arrangements.  The  glare  from  the  bare  light  sources  is 
particularly  aggravating  in  street  cars,  and  all-frosted  bulbs 
will  not  do  much  to  correct  the  fault.  Bare  carbon  lamps  of 
low  brilliancy  might  be  bearable,  but  bare  tungsten  lamps  never. 
The  car  ceilings  are  low,  and  there  is  a  vista  along  which  the  eye 
gazes.  There  are  usually  advertising  cards  to  attract  the  attention 
towards  the  upper  parts  of  the  car,  and  there  are  unavoidable 
changes  of  the  intensities  from  jarring  and  voltage  fluctuations 
that  soon  tire  the  accommodating  muscles  of  the  eye.  Hence  the 
street  railway  lamps  should  unquestionably  be  equipped  with 
shades  that  at  least  protect  the  eyes  of  the  passengers. 

Every  street  railway  must  operate  its  lighting  circuits  and  its 
power  circuits  as  one.     Hence  at  the  very  time  when  the  lights 


HIBBEN  :     STREET   RAILWAY   ILLUMINATION 


595 


are  most  needed,  the  load  on  the  system  is  the  greatest,  and  the 
fluctuations  of  voltage  are  increased  correspondingly.  This 
trouble  from  voltage  fluctuations  was  very  apparent  with  the 
use  of  carbon  filament  lamps,1  but  it  becomes  very  much  less 
troublesome  with  the  tungsten  lamps,  since  their  candle-powers 
do  not  change  so  rapidly  at  the  different  pressures.  This  can 
be  seen  in  Fig.  7.  For  instance,  if  both  types  of  lamps  gave 
their  full  rated  or  100  per  cent,  candle-power  at  500  volts,  then 
when  the  voltage  fell  to  such  a  value  as  440,  as  it  often  does  on 
street  cars,  the  tungsten  lamps  would  still  give  60  per  cent,  of  their 
initial  candle-power,  but  the  carbon  lamps  would  be  furnishing 


Tu 

ngsien 

Lamps 

Cart 

on  Lamps* 

Voltage  at  Lamps 

Fig.  7.— Voltage  candle-power  variation  curves. 

only  40  per  cent,  candle-power.  If  a  lighting  installation  is 
originally  planned  to  furnish  a  certain  adequate  illumination  at 
the  minimum  voltage,  it  can  easily  be  seen  how  much  greater 
must  be  the  investment,  and  the  waste,  in  doing  this  with  carbon 
rather  than  with  metal  filament  lamps. 

As  to  the  actual  foot-candle  values  in  a  typical  car,  note  the 
plotted  results  of  photometric  tests  that  appear  in  Figs.  8,  9,  and 
10.  Fig.  8  shows  the  transverse  and  longitudinal  distributions  in 
the  car  which  is  shown  in  Fig.  5.  Here  the  typical  differences 
between  the  bare  carbon,  and  the  shaded  tungsten  lamps  is  in- 
dicated.   In  Fig.  6  are  shown  the  test  stations  at  which  horizon- 

1  In  fact  over  35  per  cent,  of  the  operating  companies  report  considerable  trouble 
from  this  source. 


50 


TRANSACTIONS  I.  E.   S. — PART  II 


tal  foot-candles  were  measured,  at  the  reading  height  of  37 
inches.  In  Fig.  8,  the  dotted  lines  represent  the  results  from  the 
bare  carbon  lamps,  seventeen  of  which  were  burning  between 
the  bulkheads.  These  carbon  lamps  were  in  clusters,  five  above 
test  station  8-C,  four  above  each  of  the  three  stations  11-C,  5-C, 
and  2-C  (see  Fig.  6).  Compare  the  longitudinal  distributions 
along  lines  A  and  B,  obtained  from  these  seventeen  carbon  lamps, 
with  the  illumination  furnished  by  the  four  94-watt  tungsten 
lamps  with  shades,  that  were  placed  as  shown  in  Fig.  6.  The  aver- 
ages of  lines  A  and  B  with  the  bare  carbon  lamps,  are  1.9  and  2.4 


-Tungstens  and  Shades 
-Bare  Carbon  Lamps 


m 


sg 


Transverse  Section       Station    3 


Transverse  Section       Station    7 


| 

5 

,  lest  Line      B 

— " 

•>  4 

i — — * 

pN^ 

S 

^V 

.   J* 

\    \ 

o  3 

est  Line 

x.  \ 

> 

g 

\. 

■  -" 

~- 

-^ 

■*, 

""*" 

*- 





i-_  '  -"■-.' 

_. 

^ 

Line     A 

fl 

Longitudinal     Distribution  .  Along    Test  Lines    A  and  B 

Fig.  8.— Illumination  curves  from  tests  of  car  shown  in  Figs.  5  and  6. 

foot-candles  respectively.  The  corresponding  averages  with  the 
tungsten  lamps  and  shades  are  374  and  4.32.  Down  the  aisle 
(test  line  C)  the  average  with  carbon  lamps  is  2.6  and  with  the 
shaded  tungsten  lamps  4.53,  so  that  when  considering  the  whole 
car,  the  total  averages  are  2.24  for  carbon,  and  4.13  for  tungsten 
lamps  with  shades.1 

This  is  the  typical  comparison  between  bare  carbon  lamps 
and  shaded  tungsten  lamps.  The  following  table  gives  the 
detailed  story: 

1  In  averaging,  double  weight  is  given  to  lines  A  and  B,  to  arrive  at  a  true  average  for 
whole  car. 


hibben:    street  railway  illumination 


597 


TABLE  II.— Illumination  Data Bare  Carbon  vs. 

Shaded  Tungsten  Lamps. 

Bare  Carbon  Shaded  Mazda 

No.  of  lamps 170  4-° 

Candle-power  per  lamp 16.3  75-° 

Watts  per  lamp 64.0  91.0 

Total  watts 1088.0  364.0 

Total  generated  lumens 357<>.o  2952.0 

Area  car  floor 278.0  278.0 

Watts  per  square  foot    3-91  I-3I 

Average  foot-candles 2.24  4. 13 

Useful  lumens 623.0  1 150.0 

Utilization  factor 17-5%  39% 

Relative  efficiency 45  Jo  100% 

As  regards  the  illumination  when  small  bare  tungsten  lamps 
are  substituted  for  bare  carbon  lamps,  a  test  was  made  in  the 
car  of  Fig.  5,  in  which  the  arrangement  of  carbon  lamps  was  in 
groups  as  previously  mentioned,  and  which  was  the  same  as 
the  arrangement  of  the  23-watt  tungsten  lamps. 

*94  WATT  LAMPS  WITH  SHADES M  ON  CENTER-LINE  OF  CAR    & 
X  40  WATT    X  ALL-FR05TED  X  ON  SIDE  DECKS  X  OF  CAR 


FEET  2  4  |6 

STATIONS 


£3456 
Fig.  9.— Illumination  curves  from  tests  in  car  shown  in  Fig.  11. 

With  the  carbon  lamps,  the  averages  along  lines  A,  B,  and  C 
were  as  before,  namely  1.9,  2.4  and  2.6  foot-candles.  The  23-watt 
bare  tungsten  lamps  gave  corresponding  foot-candles  of  2.1,  2.6 
and  2.9.  The  total  car  average  illumination  in  the  two  cases 
was  2.24  and  2.46,  or  a  gain  of  practically  10  per  cent,  in  foot- 
candles  resulting  from  the  use  of  the  tungsten  lamps.  The  util- 
ization efficiency  of  the  bare  tungsten  lamps  was  24  per  cent. 

A  third  test  will  show  the  comparison  of  bare  all-frosted 
tungsten  lamps  vs.  clear  bulb  tungsten  lamps  with  shades.  Fig. 
9  shows  the  longitudinal  distribution  in  a  car  of  the  plan  shown 
4 


593 


TRANSACTIONS  I.  F,.   S. — PART  II 


by  Fig.  ii,  the  full  lines  representing  horizontal  foot-candles 
from  the  94-watt  lamp,  and  the  dotted  lines,  the  results  from 
the  all-frosted  bare  40-watt  lamps.  Fig.  10  shows  the  transverse 
distribution  of  these  latter  two  cases.  (Fig.  10  is  a  section  at 
station  3,  Fig.  II.) 

The  car  between  bulkheads  in  this  case  had  three  94-watt 
tungsten  lamps  on  the  center  deck  ceiling,  and  eight  40-watt 
round  bulb  tungsten  lamps  arranged  along  the  side  decks.  These 
latter  unshaded  lamps  gave  a  bad  glare,  and  were  left  in  the 


1 


..M 


40  W.  TUNGSTEN  LAMPS? 
BARE.  ALL-FROSTED   | 

i    id vA 

-i-7'-10/ 


k 


25 

o 


ABC 

Fig.  10. — Illumination  of  car  shown  in  Fig.  n,  taken  at  station  No.  3. 

car  only  to  afford  two  emergency  lighting  circuits.  The  fol- 
lowing table  gives  the  details  of  the  tests  recorded  in  Figs. 
9  and  10: 

TABLE  III. — Illumination  Data Bare  Versus 

Shaded  Tungsten  Lamps. 

Bare  Tungsten  Shaded  Tungsten 

lamps  lamps 

No.  of  lamps 8.0  3.0 

Candle-power  per  lamp 30.0  75.0 

Watts  per  lamp 39.0  84.3 

Total  watts 312.0  253.0 

Total  generated  lumens 2560.0  2142.0 

Area  car  floor 1940  194.0 

Watts  per  square  foot   1.62  1.31 

Average  foot-candles 2.71  3.24 

Useful  lumens 525.0  62S.0 

Utilization  factor 22.08%  29.30% 

Relative  efficiency   75%  100% 


HIBBEN  :     STREET    RAILWAY    ILLUMINATION 


599 


It  may  be  interesting  to  note  that  at  stations  3-A,  3-B  and 
3-C  of  Fig.  11,  the  foot-candles  on  a  plane  45  °  to  the  rear 
were  1.68,  2.34  and  2.31.  At  stations  6-A,  6-B  and  6-C  the 
values  were  1.97,  2.52  and  2.59  foot-candles. 

A  summary  of  the  above  three  cases  indicate  that  (1)  bare 
carbon  lamps  are  hardly  comparable  with  shaded  tungsten  lamps, 
being  very  inefficient.  (2)  Replacing  carbon  lamps  with  small 
bare  tungsten  lamps  is  not  advisable  and  affords  but  a  small 
gain  in  the  quantity  of  the  illumination  (greatly  increasing  the 
harshness  of  the  light),  although  considerably  reducing  the 
wattage.  (3)  Using  all-frosted  tungsten  calls  for  a  con- 
sumption of  more  energy  and  produces  much  less  useful  light 
than    shaded   tungsten   lamps.      (4)    Tungsten    lamps    properly 


^r-9Ve-8^e'-&"w-8H-r-8W-8"-*f-«-£-8:H 

*     .  ■     :  Wv.  V    ■■«»■■.  !w  ■;  Iff'    ;  !t?     .  qp — r 


iT^l  Jl    j   fTTl   [p~3  [T^  fl    5  rf^l   [|    I  £      .     select) 

:  4'    ■'  '■  +  ''     ':      i    *'■  fl     ►      i     :  *3    ;:      1 


S  SELECTOR 


o* 


-4'-0' 


-5-4}- 


•t* — 4-r-t 


■33--e- 


•  94  WATT  TUNGSTEN  LAMPS  WITH  SHADES 

*  40      •  •  ■        ALL-FROSTED 

Fig.  11. — Plan  of  street  railway  car  showing  location  of  lamps  and  test  stations. 

equipped  with  shades  will  reduce  the  wattage  19  per  cent,  as 
compared  with  all-frosted  tungsten  lamps ;  7  per  cent,  as  com- 
pared with  bare  tungsten  lamps ;  67  per  cent,  as  compared  with 
bare  carbon  lamps ;  while  simultaneously  in  the  respective  cases, 
the  shaded  tungsten  lamps  will  cause  increases  in  illumination 
of  19.5  per  cent. ;  68.0  per  cent,  and  84.5  per  cent. 

COST  COMPARISONS. 
.  No  matter  how  photometrically  efficient  a  lighting  system  may 
be  in  furnishing  the  requisite  foot-candles,  that  system  will  be 
inadvisable  in  traction  service  if  it  is  not  one  whereby  a  mone- 
tary saving  will  be  accomplished,  either  directly  by  a  saving  of 
operating  costs,  or  indirectly  through  the  medium  of  pleased 
passengers  and  greater  patronage.     The  new   shaded  tungsten 


6oo 


TRANSACTIONS  I.  E.   S. — PART  II 


filament  lamp  systems  can  accomplish  both  of  these  desired  re- 
sults. 

The  value  of  a  proportionately  small1  saving  in  energy  in 
street  railway  operation  is  seldom  given  full  consideration,  be- 
cause of  its  low  cost  of  generation.  But  energy  at  the  car  is 
more  expensive  than  power-house  costs  indicate,  and  any  sav- 
ing of  it  through  the  modern  system  of  car  lighting  will  be 
appreciable.  In  general,  power  at  a  car  will  average  about  1.5 
cents  per  kilowatt-hour.  When  a  system  of  cars  averaging 
twenty-two  64-watt  carbon  lamps  per  car  is  changed  over  to 
use  five  94-watt  tungsten  lamps,  the  energy  saved  is  938  watts, 


£ 

\ 

1 

1        1 

1 

=?-3           S 

1 

1        1 

; 

V 

1 

Ave 

age  Daily 

Peri 

d  of  Lighting 

J 

v 

-  The  time"  bctv 

-*i 

"o  "°  oa 

[-, 

H      U*     '" 
5000    .1000 

Vtar 

yAv 

erag«J 

Nearly  Average 

limi 

of  Lighting 

Ti 

11c  of 

Extingui 

»% 

4500  «  900 

^ 

4000  "^  800 

/ 

k\ 

1 

/ 

, 

\ 

.        2 

/ 

1 

k 

,*■ 

" 

/~ 

«, 

/ 

1 

^ 

lot 

llCa 

-Hours ^ 

1 

\ 

3000  ->  600 

/ 

of  iJighti 

E    . 

j 

\ 

/ 

1 

, 

" 

\ 

•3 

/ 

1 

, 

60 

1 

c 

\ 

{' 

V 

I 

1 

/ 

V" 

N 

j        / 

s 

/ 

\ 

V 

1500  Cn   300 

X 

/ 

Ti 

jl  N 

ot 

N 

/ 

/ 

1 

> 

Cara 

in  6 

mcc 

1 

t 

j    s 

/ 

1 

/ 

500  B    100 

/ 

i 

\ 

/ 

, 

\ 

00      0 

\ 

] 

2 

fooa 

I 

■ 

p 

M. 

>      1 

I)       1 

1       1 
Hi 

A.f. 

L 

1 

0     1 

1    12 
Noon 

Time  in  Hours 


Fig.  12 — Characteristic  curves  of  a  typical  street  railway  system. 


amounting  to  1.4  cents  per  car  hour.  If  there  are  1,000  cars 
on  the  system,  it  has  been  found  that  there  will  be  about 
1,500,000  car  hours  of  lighting  per  year,  on  the  average,  mak- 
ing a  total  saving  per  year  of  $21,000.  The  cost  of  the  shade 
equipment,  the  excess  of  the  cost  of  tungsten  lamps  over  car- 
bon lamps,  and  the  cleaning  expenses,  will  of  course  reduce  this 
savings  perhaps  20  per  cent. 

Approaching  the  problem  of  economics  from  a  different  angle, 
one  may  consider  Fig.  12,  which  shows  the  characteristics  of  a 

1  A  34-ton  car  with  25  passengers,  consumes  in  its  two  75  horse-power  motors  about  128 
amperes  at  500  volts. 


HIBBEN  :     STREET    RAILWAY    ILLUMINATION  6oi 

representative  street  railway  system.  In  this  figure  the  full-line 
curve  represents  the  total  number  of  cars  in  service  at  any  par- 
ticular hour  of  the  day.  The  dashed-line  curve  in  the  shape  of 
the  hyperbola  represents,  for  the  various  months  of  the  year, 
the  time  of  lighting  and  of  extinguishing  the  lamps.  The  dot- 
ted line  gives  a  summation  of  the  total  car-hours  of  lighting  for 
one  period  of  darkness,  being  based  on  a  yearly  average.  In 
tabulated  form,  a  representative  lighting  schedule  is  given  as 
follows : 


TABLE  IV.— A  Typical  Street  Railway  Lighting  Schedule 

Hours  per  Hours  per 

Months                    Period  of  Lighting  Day  Month 

January- •••   3:35  P.  M.  to  6:50  A.  m.  15:15  472 

February.  . .  4:15  p.  m.  to  6:20  a.  m.  14:05  394 

March 5:00  p.  m.  to  5:45  a.  m.  12:45  395 

April 5:50  p.  M.  to  4:50  a.  M.  11:00  330 

May 6:50  p.  M.  to  3:45  a.  M.  8:55  276 

June 7:35  p.  M.  to  3:00  A.  M.  7:25  222 

July 7:40  p.  M.  to  3:00  a.  M.  7:20  227 

August 7:00  P.  M.  tO  3:45  A.  M.  8:45  271 

September  .  5:50  p.  m.  to  4:45  a.  m.  10:55  328 

October   •  • .  4:30  p.  m.  to  5:40  a.  m.  13:10  408 

November..   3:40  p.  m.  to  6:20  A.  M.  14:40  440 

December  . .  3:15  p.  m.  to  6:55  a.  m.  15:40  486 

^Average-  •  5:25  p.  m.  to  5:05  a.  m.  11:37  354 


When  considering  the  average  number  of  cars  in  service  dur- 
ing each  hour  of  the  dark  period,  and  multiplying  this  number  by 
the  hours  each  car  is  lighted,  it  is  seen  that,  for  each  day,  there 
will  be  approximately  3,900  car-hours  of  lighting  on  this  1,000- 
car  system.  This  amounts  to  1,500,000  car-hours  per  year. 
Now  a  careful  analysis  made  by  other  authorities,  of  power 
costs  of  this  representative  system,  based  on  the  wattages  of  the 
old  carbon  lamp  and  the  new  tungsten  filament  lamp  and  shade 
equipment,  show  a  total  cost  of  lighting  power  per  year  of 
$35,500  and  $14,150  respectively.  Therefore  a  summary  of 
lighting  costs  will  be  about  as  follows : 

1  In  southwestern  United  States,  and  in  agricultural  or  similar  regions  having  much 
clear  weather  the  average  hours  per  month  falls  as  low  as  225. 


602  TRANSACTIONS    I.    E.    S. — PART    II 

TABLE  V.— Lighting  Costs,  With  the  New  and  the  Old  Systems. 

Carbon  lamps  Tungsten  lamps 

Car-hours  of  lighting  per  year 1,500,000  1,500,000 

1  Average  number  of  lamps  per  car- ■  •        22.38  5-94  watt  4.48-23  watt 
Lamp  hours  of  lighting  per  year   . ...   33,600,000  7,500,000  6,730,000 

Average  life  of  lamp  in  hours 1,800            2,000  2,000 

Yearly  number  of  lamp  renewals 18,650            3,75°  3>365 

Net  cost  of  lamps  (on  $2,500  contract)   $2,238.00  $2,396.25  $955-65 

$  3,352.00 
Yearly  cost  of  power  for  lighting $35,500.00  14,150.00 

2  Estimated  expense  of  shade   equip- 

ment    —  2,000.00 


Final  total  yearly  cost $37, 738.00  $19,502.00 

The  above  figures,  showing  a  probable  saving  of  $18,000.00 
or  $18.00  per  car  per  year,  compare  closely  with  the  estimate  as 
previously  presented,  and  moreover  they  represent  a  very  con- 
clusive argument  for  the  adoption  of  the  modern  lighting  sys- 
tem. This  is  but  the  financial  viewpoint,  and  one  should  not 
loose  sight  of  the  accompanying  increase  in  the  amount  of  light 
(84  per  cent,  increase,  see  table  11)  nor  of  the  better  qualities 
of  the  illumination. 

SUMMARY  OF  THE  STATUS  AND  PROSPECTS  OF 
CAR  LIGHTING. 

Briefly,  the  most  important  considerations  to  the  gathered  from 
the  above  discussions  are  summarized  as  follows: 

The  bare  carbon  or  graphitized  filament  lamps  are  no  longer 
advisable  nor  economical  for  street  railway  service,  on  account 
of  short  life,  high  energy  consumption,  poor  quality  and  unsteadi- 
ness of  light,  and  the  wiring  costs  incidental  to  the  installing  of 
a  large  quantity  of  them. 

Small  bare  tungsten  lamps  replacing  them  are  nearly  as  waste- 
ful from  the  standpoint  of  not  having  their  light  directed  by 
proper  shades,  and  in  addition  are  extremely  dazzling  and 
optically  harmful. 

Large  tungsten  lamps,  of  94  or  56-watt  sizes,  seem  the  most 

1  The  fractions  of  lamps  come  from  an  average  of  motor-cars  and  of  trailers,  the  latter 
having  fewer  lamps,  and  in  the  case  of  the  tungsten  lamps,  having  but  the  one  circuit  of 
five  94-watt  units. 

2Basedon  (1)  First  cost  $7, coo,  or  interest  of  $2So.oo.  (2)  10  per  cent,  breakage  and 
depreciation,  or  $700.00.  (3)  Cleaning  twice  monthly  $1.05  per  year  per  car  or  $1,050.00.  As 
regards  this  cleaning  expense,  the  data  is  meager.  A  fair  figure  for  dusting  once  daily 
seems  to  be  60^  per  car  per  month,  and  washing  twice  monthly,  6o#  to  $1.25  per  car  per 
month. 


HIBBEN  :     STREET    RAILWAY    ILLUMINATION  603 

economical,  in  conjunction  with  shades.  These  units  may  more 
than  double  the  previous  (most  generally  inadequate)  values  of 
illumination,  at  one-third  of  the  energy  cost,  giving  a  soft  dif- 
fused light  and  forming  a  neat   decorative  unit. 

The  candle-power  of  the  tungsten  lamps  is  much  more  steady 
and   less   affected   by   voltage   fluctuations. 

The  net  cost  of  lighting  with  the  new  units  is  much  less  than 
when  carbon  lamps  are  used;  the  cost  of  the  tungsten  lamps  is 
slightly  greater,  but  this  is  more  than  compensated  for  by  the 
energy  saved. 

All  these  items,  and  others  that  appear  from  a  study  of  this 
subject,  seem  to  indicate  a  prospect — almost  a  certainty — of  a 
rapid  and  a  large  development  in  street  railway  car  lighting. 
Up  to  the  present  time  there  have  been  perhaps  a  half  dozen 
installations  of  large  tungsten  units,  and  shades.  Bare  tungsten 
(together  with  a  few  tantalum)  lamps  are  in  service  on  approxi- 
mately 28  per  cent.,  bare  graphatized  lamps  on  10  per  cent.,  and 
bare  carbon  lamps  on  60  per  cent,  of  the  175,000  to  200,000  cars 
in  service  in  the  United  States. 

Very  soon  the  traveling  public  may  be  expected  to  demand 
better  lighted  cars.  As  the  passenger  looks  lengthwise  of  the  car 
(75  per  cent  have  cross  seats)  or  up  at  the  advertising  cards  (on 
95  per  cent,  of  all  cars)  he  cannot  help  but  be  affected  by  the 
glare  from  the  old  lighting  systems.  If  he  may  be  freed  from 
this  annoyance,  and  does  not  have  to  strain  his  eyes  because 
of  insufficient  or  fluctuating  illumination,  surely  he  becomes  a 
more  valuable  asset  to  any  railway  company. 

Therefore  this  modern  lighting  system  seems  destined  to  be 
mutually  advantageous  to  street  railway  companies  and  passen- 
gers. It  is  hoped  that  further  investigations  of  this  subject 
will  soon  afford  valuable  additions  to  the  somewhat  preliminary 
study  that  this  paper  presents. 


604  TRANSACTIONS   I.    E.    S. — PART    II 

DISCUSSION. 

Mr.  L.  C.  Porter:  A  little  over  two  years  ago  Air.  Stickney 
took  up  the  question  of  street  car  lighting.  We  went  to  several 
railway  companies  to  put  the  proposition  up  to  them.  The  first 
question  that  arose  was :  Will  the  lamps  stand  the  severe  use 
of  street  car  service?  At  that  time  we  had  results  of  tests  which 
had  been  conducted  on  some  of  the  United  States  battleships 
where  tungsten  lamps  had  gone  through  target  practise,  full 
power  run  and  several  other  severe  tests,  giving  very  good  re- 
sults. We  also  had  tests  on  ferry  boats  where  the  lamps  oper- 
ated satisfactorily. 

With  the  co-operation  of  Mr.  Hoist  of  the  Bay  State  Street 
Railway  Company,  we  equipped  some  of  their  cars  with  56-watt 
tungsten  lamps  and  prismatic  reflectors.  The  question  of  the 
different  sized  lamps  was  studied  and  it  seemed  best  to  use 
the  56-watt  lamps  for  several  reasons.  The  principal  reason 
was  that  with  the  94-watt  lamp  the  failure  of  one  lamp  would 
put  the  car  in  darkness  for  at  least  a  few  minutes  until  the  con- 
ductor could  find  the  selector  switch  and  operate  it;  while  with 
the  56-watt  lamp  there  were  two  circuits,  and  since  the  lamps 
are  wired  on  alternate  circuits  the  failure  of  one  lamp  would 
not  plunge  the  car  into  darkness.  Another  advantage  which  was 
found  in  using  the  56-watt  lamps  was  that  the  lamps  could  be. 
spaced  a  little  closer  together,  giving  less  sharp  shadows  than 
are  likely  to  maintain  in  the  94-watt  system.  For  instance,  look 
at  the  diagram  on  page  598;  and  consider  a  passenger  seated  in 
the  second  seat  from  the  left-hand  end  of  the  car,  holding  his 
paper  on  what  is  an  average  reading  plane;  that  is,  450  (3  feet 
above  the  floor).  When  facing  the  right-end  of  the  car  the 
passenger  will  get  very  good  light  on  the  paper,  but  when  he  faces 
the  left-end  the  illumination  on  the  paper  will  not  be  so  good, 
because  the  light  which  is  then  shining  on  it  will  come  from  a 
considerable  distance;  whereas  if  a  lamp  were  placed  halfway 
between,  there  would  be  practically  the  same  amount  of  illumin- 
ation no  matter  whether  the  reader  were  facing  forward  or 
backward. 

Twenty-eight-foot  cars  and  34-foot  cars  were  equipped  with 
56-watt  lamps.    In  installing  this  system  we  spaced  the  lamps  6 


STREET   RAILWAY   ILLUMINATION  605 

feet  apart,  in  this  way  bringing  the  two  end  lamps  2  feet  from 
the  ends  of  the  car,  with  another  lamp  on  each  platform,  one 
in  the  headlight  and  one  56-watt  lamp  in  each  sign  of  the  car. 
Where  two  lamps  had  been  used  per  sign,  we  installed  2  28- 
watt,  half  voltage  lamps  (equivalent  to  one  56-watt,  full  voltage 
lamp),  thus  making  two  full  circuits  of  5  56-watt  lamps  each. 

Photometer  tests  were  conducted  in  cars  so  equipped  and  it 
was  found  that  in  a  car  that  had  previously  been  equipped  with 
25-watt  64  carbon  lamps,  consuming  1,600  watts,  we  reduced 
this  consumption  to  560  watts  by  using  the  56-watt  lamps  which 
gave  approximately  3  effective  lumens  per  watt;  whereas  only 
0.4  was  obtained  with  the  carbon  lamps. 

Both  intensive  and  extensive  type  reflectors  were  tried.  It  was 
found  that  the  intensive  type  of  reflector  gave  the  better  results. 
It  gave  a  better  average  intensity  and  threw  the  maximum 
amount  of  light  where  it  was  most  needed,  that  is,  over  the 
seats. 

An  interesting  shadow  test  was  conducted  in  which  Mr.  Hoist 
seated  himself  in  various  seats  in  the  car  and  was  surrounded 
by  a  number  of  men,  to  see  if  they  could  cast  objectionable 
shadows  over  the  paper  he  was  reading.  It  was  found  impos- 
sible to  do  this. 

Six  of  these  cars  were  operated  by  the  Bay  State  Company  over 
various  conditions  of  roadbed,  continuously  for  over  a  year  in 
order  to  make  absolutely  sure  that  both  the  reflectors  and  the 
lamps  would  stand  the  service.  At  the  end  of  the  year  they  had 
proved  very  satisfactory.  Those  tests  were  reported  by  Mr. 
Hoist  in  the  Electric  Railway  Journal  for  September,  1912. 
As  a  result  of  the  tests  the  Bay  State  Street  Railway  Company 
is  equipping  all  of  its  new  cars  with  this  system  of  lighting  and 
also  rewiring  some  of  the  old  ones.  There  are  several  other 
companies  throughout  the  country  that  are  doing  likewise,  prov- 
ing that  this  system  is  a  paying  proposition. 

Mr.  E.  B.  Rowl  :  Mr.  Hibben's  paper  is  an  excellent  one 
for  our  Transactions  because  it  adds  something  new.  Street 
car  lighting,  or  rather  the  proper  illumination  of  street  cars,  is  a 
relatively  new  development  and  there  are  tremendous  possibilities 
in  the  way  of  improvements. 


606  TRANSACTIONS   I.    E-    S. — PART    II 

During  Mr.  Porter's  discussion  it  occurred  to  me  that  circum- 
stances may  alter  cases :  this  should  be  remembered  in  discus- 
sions of  the  56-watt  2-circuit  system  versus  94- watt  1 -circuit 
system,  from  the  standpoint  of  lamp  failures.  The  consequences 
of  the  total  though  brief  loss  of  light  in  a  car  which  would  result 
when  a  lamp  fails  in  the  latter  system  would  depend  to  a  large 
extent  on  whether  it  is  totally  dark  outside  of  the  car  and  con- 
sequently on  the  territory  over  which  the  car  operates.  In  most 
city  streets  there  is  enough  external  light  coming  in  so  that  the 
total  failure  of  the  light  in  the  car  is  not  of  so  much  importance, 
whereas  in  running  over  a  private  right  of  way  or  over  country 
roads  it  is  of  great  importance. 

Again,  as  regards  ability  to  read  with  the  one-circuit  and  two- 
circuit  systems, — that  may  depend  on  the  type  of  car  construc- 
tion. If  it  is  a  car  having  the  passengers  all  facing  in  one  direc- 
tion and  not  across  the  car,  the  need  for  good  reading  illumina- 
tion, both  in  intensity  and  proper  direction,  is  greater  than  in  that 
type  of  car  which  is  rapidly  coming  into  favor  for  con  jested  city 
service.  My  point  is  that  what  has  proved  to  be  the  best  type 
of  lighting  system  for  one  installation  is  not  necessarily  the  best 
type  to  adopt  for  some  other  traction  system. 

In  Mr.  Hibben's  paper  he  mentions  in  his  introduction  that  the 
first  installation  of  lamps  with  individual  reflectors  in  street  cars 
was  in  1909  in  Dayton.  I  believe  there  were  several  isolated 
instances  where  single  carbon  lamps  were  equipped  with  pris- 
matic reflectors  prior  to  that  date.  There  is  one  such  installation 
on  the  Washington  &  Baltimore  Traction  System. 

During  the  past  few  years  or  even  months,  to  one  who  has 
been  following  the  situation,  it  is  quite  remarkable  to  note  the 
change  in  attitude  regarding  this  question  of  car  lighting.  Sev- 
eral years  ago  the  question  of  changing  the  lighting  in  the  subway 
cars  of  New  York  came  up  under  the  direction  of  the  Public 
Service  Commission.  Extensive  tests  were  made  to  show  what 
improvement  could  be  effected  by  change  in  equipment.  These 
tests  brought  out  clearly  of  course  the  advantages  from  the  stand- 
point of  uniformity  and  higher  intensity  of  using  reflectors  over 
the  lamps  and  by  substituting  tungsten  lamps  for  the  carbon 


STREET   RAILWAY    ILLUMINATION  607 

lamps  then  in  use.  The  danger  from  glare,  however,  was  not  so 
well  understood  at  that  time  and  the  use  of  reflectors  was  tem- 
porarily decided  against  because,  as  one  of  the  managers  of  the 
subway  company  expressed  it,  "the  public  was  not  educated  to 
the  need  of  reflectors."  This  is  an  attitude  which  will  probably 
always  require  a  great  deal  of  effort  to  overcome.  Something 
really  desirable  should  not  be  turned  down  because  the  public 
has  not  been  "educated  up  to  it."  I  believe  the  change  should  be 
made  and  the  advantages  proved  in  actual  practise. 

The  use  of  a  spare  lamp  with  the  selector  switch  gives  a  very 
satisfactory  system  of  illumination.  On  the  Cleveland  cars  the 
newer  ones  have  the  extra  lamp  placed  in  a  similar  fitting  to  the 
lamp  in  service  over  the  rear  platform,  these  two  lamps  being 
placed  close  together  in  individual  porcelain  enameled  bowl-shaped 
reflectors  which  are  counter  sunk  in  the  ceiling  of  the  car.  These 
reflectors  with  the  prismatic  reflectors  used  in  the  body  of  the 
car  provide  a  very  satisfactory  system  giving  practically  no  glare 
effect;  and  the  steps,  platform  and  seats  are  amply  lighted. 
As  these  are  pay-enter  cars  the  selector  switch  is  always  within 
easy  reach  of  the  conductor  and  only  a  few  seconds  are  required 
to  locate  the  dead  lamp,  in  case  of  a  failure,  and  replace  it  by 
the  spare  lamp  at  the  first  convenient  opportunity. 

The  effect  of  changing  from  carbon  to  tungsten  lamps  from 
the  standpoint  of  change  in  intensity  with  voltage  variations  is, 
I  believe,  quite  noticeable  to  anyone  who  has  used  a  line  fre- 
quently before  and  after  such  a  change.  On  the  Columbus- 
Zanesville  Interurban  Line  I  had  an  opportunity  of  observing 
this  fact  and  while  I  knew  what  to  expect  I  was  actually  as- 
tounded by  the  improvement  after  the  substitution  of  the  tung- 
sten lamps.  Before  the  change  when  a  car  pulled  out  of  the  Co- 
lumbus station  it  was  necessary  to  stop  reading,  roll  up  one's 
coat  for  a  pillow  and  try  to  get  some  sleep ;  the  voltage  dropped 
so  much  that  the  car  was  almost  in  darkness.  After  the  tungsten 
lamps  were  put  in  one  could  read  fairly  comfortably  during  the 
entire  trip,  even  though  the  voltage  fluctuation  on  that  line  is 
quite  extreme. 

The  use  of  auxiliary  emergency  lighting  might  be  mentioned 
because  this  ought  sometimes  to  be  provided  on  lines  where  an 


608  TRANSACTIONS   I.   E.    S. — PART    11 

absolute  or  extended  failure  of  the  service  might  result  in  panics 
and  serious  accidents.  The  use  of  a  small  storage  battery  equip- 
ment with  a  few  lamps  placed  at  convenient  points  in  the  car 
will  provide  light  instantly  when  other  sources  fail. 

It  may  be  interesting  to  the  I.  E.  S.  members  to  follow  a  series 
of  tests  in  a  study  of  coach  lighting  which  has  just  been  conducted 
under  the  auspices  of  the  Railway  Electrical  Engineers.  Al- 
though these  tests  were  conducted  with  primary  reference  to 
steam  road  conditions,  the  illumination  requirements  in  coaches 
are  so  similar  to  those  in  street  and  interurban  electric  cars  that 
these  tests  should  be  of  considerable  help  in  designing  the  light- 
ing systems  for  electric  cars.  These  tests  were  quite  extensive, 
involving  a  consideration  of  shadow  effects,  etc.,  and  will  un- 
doubtedly be  published  in  full  shortly. 

Mr.  G.  H.  Stickney  :  I  believe  thoroughly  that  the  use  of  re- 
flectors with  tungsten  filament  lamps  in  street  car  lighting  is  the 
coming  practise,  both  on  account  of  the  superior  lighting  effect 
and  economy.  (The  installation  of  such  units  on  the  Bay  State 
Street  Railways,  described  in  the  Railway  Electrical  Engineer 
of  September  28th  191 2,  was  the  first  of  the  recent  installations 
of  this  type.)  On  the  other  hand,  I  cannot  quite  agree  with  the 
author  that  the  use  of  the  bate  tungsten  filament  lamp  is  not  justi- 
fied, at  least  as  a  temporary  expedient,  for  certain  conditions.  In 
the  first  place,  on  account  of  the  expense  and  time  required  to  re- 
wire and  equip  old  cars  already  in  active  service,  it  is  impracticable 
for  many  roads  to  change  over  completely  immediately.  In  the 
second  place,  the  demand  for  soft,  diffused  light  in  street  cars 
is  nowhere  near  as  great  as  in  train  lighting  or  other  interior 
lighting  installations,  since  the  passengers  are  not  in  the  cars  for 
relatively  long  periods  of  time,  and,  wearing  hats,  they  do  not 
have  their  eyes  exposed  to  the  glare  to  the  same  extent  as  under 
these  other  conditions.  I  have  ridden  in  the  subways  in  New 
York,  which,  as  you  know,  are  equipped  with  bare  tungsten  fila- 
ment lamps,  and  it  has  been  my  observation  that  relatively  few 
of  the  passengers  have  been  conscious  of  the  glare  effect,  while 
many  have  appreciated  the  higher  intensity  provided  and  the 
improved  steadiness  of  the  light  under  voltage  variation,  jwhicli 


STREET    RAILWAY    ILLUMINATION  OCX) 

is  a  very  decided  advantage  introduced  by  the  tungsten  filament 
lamps,  whether  with  or  without  reflectors. 

I  firmly  believe  that  the  question  of  economy  alone  should 
induce  the  railways  to  immediately  adopt  reflectors  in  all  new 
cars,  and  that  in  the  long  run  it  will  compel  their  adoption  even 
at  the  expense  of  re-wiring,  for  all  roads  in  which  costs  are 
carefully  figured. 

Mr.  Porter  has  mentioned  tests  in  which  we  are  assisting  the 
New  York  Municipal  Railways.  I  hope  that  it  will  be  possible 
to  make  these  tests  public,  as  they  are  probably  some  of  the  most 
complete  yet  undertaken.  Practically  every  arrangement  of  light- 
ing that  promises  to  be  suitable  for  the  subway  cars  in  question 
is  being  tried  out  and  observed. 

Referring  again  to  the  paper,  the  author  states  that  there  are 
half  a  dozen  lines  using  tungsten  filament  lamps.  I  think  this 
is  a  little  conservative,  as  I  am  sure  we  have  handled  a  consider- 
ably larger  number  than  that  in  our  own  office. 

Mr.  V.  R.  Lansingh  :  The  substitution  of  tungsten  lamps 
for  carbon  lamps  naturally  increases  eyestrain,  and  it  is  a 
question  with  me  whether  or  not  the  economy  gained  by  the  use 
of  the  new  lamps  and  the  increase  in  illumination  is  not  more 
than  offset  by  the  decrease  in  eye  comfort.  Furthermore,  when 
the  railroad  has  made  such  substitution  the  economy  to  be  gained 
by  the  use  of  reflectors  is  smaller  than  it  would  have  been  if 
they  had  started  in  at  the  beginning  and  changed  the  entire  sys- 
tem. It  is  a  question,  when  the  new  lamps  have  once  been  in- 
stalled whether  there  would  ever  be  a  change  to  the  old  wiring 
system  and  in  new  lamps  and  reflectors  installed.  I  believe  I 
would  rather  see  the  old  system  remain  in  place  and  the  new 
complete  system  will  therefore  come  sooner. 

Mr.  C.  W.  Bettcher:  I  believe  that  the  fluctuation  in  the 
candle-power  of  carbon  lamps,  due  to  the  change  in  voltage,  is 
much  more  objectionable  than  the  glare  from  the  tungsten  fila- 
ment lamps  when  they  are  installed.  Most  people  do  not  care  so 
much  about  the  glare;  in  fact  they  do  not  notice  it  when  reading; 
but  certainly  the  change  in  candle-power  is  very  objectionable, 
especially  where  a  good  many  stops  are  made  and  frequent  start- 
ing causes  a  temporary  drop  in  voltage. 


6lO  TRANSACTIONS    I.    E.    S. PART    II 

Mr.  Ward  Harrison  :  In  comparing  the  relative  advantages 
of  the  56  and  94-watt  lamps,  the  type  of  car  should  be  considered. 
Since  in  Cleveland  pay-enter  cars  are  used  almost  exclusively, 
the  short  circuiting  switch  is  always  within  easy  reach  of  the 
conductor  stationed  at  the  fare  box  and  a  burn-out  in  the  car 
will  not  cause  the  lamps  to  be  extinguished  for  more  than  a  few 
seconds.  If  on  the  other  hand  the  type  of  car  is  such  that  the 
conductor  is  not  always  at  the  same  place,  there  is  surely  an 
advantage  in  having  two  circuits. 

Mr.  R.  B.  Ely:  Within  the  past  two  or  three  years  a  num- 
ber of  papers  have  been  presented  on  car  lighting.  One  of  these 
has  advocated  a  system  by  which  the  light  would  be  directed 
downward  and  forward,  the  light  source  being  semi-concealed. 
Such  a  system  would  be  an  improvement  over  the  present  method 
of  lighting  the  near-side  car.  New  cars  of  this  type  are  fully 
equipped  with  illuminated  signs,  illuminated  steps,  etc.  The 
lighting  installation  consists  of  bare  lamps  placed  along  both 
sides  over  the  seats.  A  semi-concealed  system  of  lighting  could 
possibly  be  used  to  advantage  in  this  type  of  car. 

Mr.  J.  B.  Jackson  :  There  is  one  item  that  I  believe  should 
be  brought  out  more  clearly  in  a  cost  analysis  and  that  is  first 
cost  of  equipment.  This  is  a  point  which  is  of  great  importance 
to  the  street  railway  company  and  one  which  will  be  the  prin- 
cipal consideration  in  changing  the  lighting  equipment  in  present 
cars.  It  seems  as  though  Mr.  Hibben's  cost  of  $7.00  per  car  is 
rather  low  as  that  would  indicate  a  cost  of  $1.17  per  unit  for  the 
six-unit  equipment.  I  believe  with  the  special  holders  required, 
the  simplest  equipment  possible,  i.  e.  holder,  socket  and  reflector, 
could  not  be  installed  for  much  less  than  $2.50  per  unit.  This 
will  make  a  slight  increase  in  the  last  item  of  the  cost  analysis 
Table  V.  I  would  also  suggest  that  the  words  "per  year"  be 
added  to  the  item  making  it  read  "estimated  expense  of  shade 
equipment  per  year." 

Mr.  S.  G.  HibbKn  (In  reply)  :  Concerning  shadows,  re- 
ferred to  by  Mr.  Porter,  I  have  in  mind  one  case  where  the  ob- 
jection to  lighting  with  the  center-deck  units  was  put  forth  with 
the  effect  that  a  standing  passenger  would  shield  a  seated  pas- 


STREET    RAILWAY    ILLUMINATION  6ll 

senger,  particularly  if  the  passenger  were  seated  on  a  longitudinal 
seat.  In  a  brief  investigation  I  found  that  there  are  ordinarily 
no  grounds  for  such  an  objection.  The  light  from  the  units 
at  one  side  or  the  other  of  the  standing  passenger  will  give  suffi- 
cient illumination  so  that  there  will  be  no  sharp  shadows  on  the 
seated  passengers. 

Formerly  the  proposition  has  been  advanced  that  center  deck 
units  would  not  light  the  advertising  cards.  In  actual  practise 
the  center  of  the  light  sources,  if  this  type  of  unit  were  used, 
(referring  to  the  fixture  that  projects  entirely  below  the  deck) 
would  be  6  or  7  inches  below  the  car  ceiling,  and  there  would 
be  full  illumination  of  all  sides.  In  fact  in  all  cases  that 
I  have  seen,  one  is  better  able  to  read  these  cards,  and  with  more 
comfort.  One  can  hardly  see  at  all  the  sign  that  is  directly  be- 
hind a  bare  lamp  on  the  edge  of  the  half  deck. 

Mr.  Harrison  brought  out  the  point  that  on  certain  cars  per- 
haps the  conductor  would  not  be  within  reach  of  the  selector 
switch.  In  that  case  one  solution  might  be  to  place  the  switch 
in  the  motorman's  end  of  the  car  (if  it  were  not  of  the  double- 
end  type)  so  the  motorman  would  always  be  able  to  quite  con- 
veniently turn  it. 

Mr.  Ely  brought  out  the  fact,  and  a  rather  surprising  fact  it 
is,  that  a  large  number  of  cars  have  improvements  in  heating 
and  ventilation  and  illuminated  signs  and  headlights,  while  as 
yet  not  much  attention  has  been  given  to  improving  the  illumina- 
tion between  bulkheads. 

I  will  not  enter  into  any  lengthy  discussion  with  some  of 
these  lamp  men,  for  I  believe  they  will  themselves  wish  to  re- 
consider or  qualify  the  remarks  as  to  the  use  of  unshaded  lamps. 
I  advocate  reflectors  because  in  the  first  place  I  believe  that 
through  their  use  they  will  pay  for  themselves.  It  is  purely  a 
matter  of  economics,  because  if  a  reflector  can  increase  the  use- 
ful light,  as  it  certainly  can  in  most,  cases,  it  follows  that  the 
wattage  may  be  reduced,  and  that  brings  about  a  saving  that 
will  more  than  pay  for  the  added  expense  of  the  reflectors  and 
their  maintenance. 

Regarding  Mr.  Jackson's  statement  concerning  the  costs  of 
units, — possibly  these  figures  of  the  first  cost  of  $7,000 — about 


6l2  TRANSACTIONS   I.    E.    S. — PART   II 

$7.00  per  car — are  a  little  bit  low  at  the  present  prices  of  these 
units.  The  price  of  these  will  drop  somewhat,  as  their  use  is 
extended.  The  cost  will  run  about  $1.17  per  unit  for  the  fixture 
and  shade.  I  do  not  consider  the  installation  cost  a  factor,  since 
the  wiring  will  be  about  as  expensive  for  the  large  number  of 
the  old  units,  as  for  the  less  number  of  new  units. 


ELY:     CHURCH  LIGHTING  613 

CHURCH  LIGHTING.* 


BY  ROBERT  B.   ELY. 


Synopsis:  This  paper  discusses  some  of  the  problems  encountered  in  the 
lighting  of  churches.  It  outlines  briefly  some  requirements  of  certain  parts 
of  churches  of  different  denominations.  Views  of  typical  interiors  are  given. 

I.  STRUCTURE. 

In  discussing  the  illumination  of  churches  it  may  be  apropos 
to  give  a  few  facts  concerning  the  history  of  the  early  ecclesias- 
tical structures  as  places  of  Christian  worship. 

The  structures  were  not  copied  from  either  the  Heathen  or 
Jewish  temples,  but  from  a  combination  hall  of  justice  and  a 
market  place,  which  was  called  a  basilica  by  the  ancients.  The 
rites  of  heathendom  were  performed  almost  exclusively  by  the 
priest,  and  the  temples  were  lighted  only  by  the  daylight  that 
came  through  the  doorways  or  interior  courts,  or  by  a  few  lamps 
that  burned  around  the  image  of  the  God.  The  temple  was  not 
regarded  as  an  assembly  room  for  worshippers,  but  only  as  an 
abode  of  the  God.  Thus  the  dark  and  mysterious  temples  were 
unsuited  for  religious  services,  in  which  the  people  were  to 
participate.  Although  the  basilica  served  its  purpose  as  a 
place  of  worship  there  was  little  or  no  significance  in  the 
structure  to  awaken  Christian  sentiment.  The  Christians  from 
an  early  period  used  the  cross  as  a  sacred  emblem,  and  in  their 
endeavor  to  indicate  their  allegiance  to  the  author  of  their  salva- 
tion they  modified  their  structures  to  the  form  of  the  cross ;  both 
the  Latin  and  Greek  crosses  were  followed.  In  either  case  the 
arms  at  right  angles  and  directly  opposite  each  other,  cut  it 
across,  and  were  given  the  name  of  transepts.  Over  the  point 
of  intersection  of  the  transepts,  the  body  of  the  cross,  a  central 
tower  or  spire  was  frequently  erected.  Beyond  the  galilee  or 
entrance  chapel,  or  in  some  instances  the  entrance  door  to  the 
transepts,  was  the  nave.  If  there  were  no  transepts  the  nave 
would  extend  from  the  choir  to  the  principal  entrances,  but 
would  not  include  the  aisles.  Side  aisles  frequently  continued 
along  the  transepts,  thus  running  around  the  whole  church ; 
sometimes  there  were  double  aisles  to  the  nave.     Beyond  the 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 

5 


614  TRANSACTIONS    I.    E.    S. — PART    II 

transepts  was  the  chapel  or  chancel,  in  which  was  situated  the 
altar;  sometimes  there  were  several  altars.  Side  chapels  will 
sometimes  be  found  on  the  side  aisles. 

The  early  Christian  churches  were  lighted  by  daylight  through 
the  construction  of  a  clear  story. 

MODERN  CHURCHES. 

There  are  scarcely  two  churches  alike  in  structure  although 
there  is  often  a  similarity  of  the  plans  of  churches  of  the  same 
denomination.  For  example,  the  Roman  Catholic  Church  has 
generally  adopted  the  Italian  Renaissance  and  the  Episcopal  the 
English  Gothic.  But  there  are  comparatively  few  classic  ex- 
amples in  existence ;  most  of  the  churches  of  the  present  day  are 
modern  adaptations  of  several  styles  of  architecture.  The  material 
used  in  building  construction  and  the  building  law  requirements 
have  made  it  necessary  to  depart  from  the  purely  classic  styles 
in  order  to  provide  for  fire-proofing  and  other  construction 
details ;  so  that  one  no  longer  sees  a  style  or  order  in  its  true  pro- 
portions. The  endeavors  of  an  architect,  therefore,  are  not 
strictly  along  the  lines  of  what  has  gone  before.  He  usually 
attempts  to  work  the  adaptation  of  one  of  the  various  styles  of 
architecture  into  a  pleasing  ensemble. 

Any  building  set  apart  for  religious  services  is  termed  a  church, 
excepting  those  buildings  of  smaller  dimensions,  which  are  called 
chapels. 

The  Gothic  style  of  architecture  predominates  in  the  construc- 
tion of  present  day  churches.  This  style  is  considered  one  of 
the  noblest  and  most  complete  in  architectural  design.  Its 
distinctive  features  are  the  Gothic  pointed  arch,  the  tendency 
toward  vertical  lines,  deep  mouldings  on  columns,  capitals,  etc., 
and  decorations  derived  from  various  kinds  of  foliage.  The 
towers  are  frequently  square  at  the  base  and  terminate  with 
lofty  spires  richly  decorated.  The  hammer  beams  and  pendants 
are  also  among  the  chief  features.  But  in  general  the  modern 
churches  are  adaptations  of  the  French  and  Italian  Renaissance. 

Intensity  of  Light. — In  considering  the  quantitative  values  of 
the  illumination  in  the  church,  experience  has  shown  that  it  is 
inadvisable  to  stipulate  a  certain  intensity  of  light  for  the  audi- 
torium, the  sanctuary  or  chancel.     It  is  advisable,  however,  to 


£>H 


Fig.  i. — Direct  illumination  of  a  church. 


Fig.  2.—  Overhead  illumination  in  a  synagogue. 


Fig.  3. — Cove  lighting  in  a  church. 


Fig.  4.— Indirect  lighting  in  a  church. 


Ely:    church  lighting  615 

determine  a  relative  ratio  between  the  intensity  of  light  of  the 
chancel,  and  that  of  the  main  part  of  the  church.  That  ratio  in  a 
number  of  effectively  illuminated  churches  is  about  2  to  1  or 
greater  in  favor  of  higher  illumination  in  the  chancel  or  sanctuary. 

Owing  to  the  general  systems  of  control,  particularly  of  electric 
lighting  installations  and  the  manipulation  of  this  control,  prior 
to  the  beginning  of  services  one-half  of  the  installation  is  usually 
in  use.  During  the  general  service,  in  which  the  congregation 
takes  part,  the  entire  chancel  and  auditorium  is  illuminated. 
While  the  sermon  is  being  delivered  the  illumination  in  the  audi- 
torium is  usually  reduced,  and  the  chancel  or  such  lamps  that  are 
used  to  illuminate  the  pastor  and  pulpit  are  used.  The  entire 
equipment  is  again  used  during  the  closing  services.  Provision 
should  be  made  in  Roman  Catholic  churches  for  illumination  of 
the  stations  of  the  cross  throughout  the  day  and  evening. 

Any  set  calculation  relative  to  watts  per  square  foot  is  inad- 
visable, owing  to  the  numerous  variables  found  in  church  struc- 
ture. Some  churches  are  illuminated  with  an  energy  consump- 
tion of  0.3  of  a  watt  per  square  foot  of  floor  area,  while  others 
require  as  high  as  2.5  watts  per  square  foot,  both  installations 
being  considered  good  examples. 

There  is,  no  doubt,  greater  intensity  of  illumination  in  the 
newer  and  reconstructed  installations  of  churches ;  yet  the  varia- 
tion in  intensity  of  illumination  is  comparatively  wide. 

Commercial  factors  such  as  costs  of  installation  and  operation 
present  themselves  in  all  but  a  few  instances,  and  tend  to  deter- 
mine the  character  and  intensity  of  the  installation  to  a  great 
extent.  A  very  elaborate  lighting  equipment  can  be  designed, 
but  unless  the  bearing  of  the  commercial  factors  of  the  case  has 
been  determined,  the  chances  are  that  the  plan  will  be  discarded. 

In  the  illuminating  engineering  work  of  the  Philadelphia 
Electric  Company  a  very  broad  policy  permits  the  lighting 
specialist  to  draw  up  plans  and  specifications  that  are  guided 
largely  in  each  case  by  the  church .  officials,  or  architect.  The 
company  aims  to  present  a  proposition  that  will  be  in  keep- 
ing with  the  architecture,  effective  in  results  from  an  illumi- 
nation stand-point,  and  economical  in  operation.  The  specialist 
treats  the  proposition  in  an  unbiased  manner.  He  is  permitted 
to  specify  any  system,  shades,  reflectors  or  reflecting  devices, 


6l6  TRANSACTIONS   I.    %    S. — PART   II 

which  in  his  judgment  will  meet  the  conditions  of  a  given  case. 
Due  attention  is  also  paid  to  possible  emergency  lighting  by  gas 
units. 

It  may  be  of  interest  to  mention  some  of  the  demands  and 
tendencies  of  the  clergymen  and  architects  in  such  cases.  In 
Philadelphia  many  architects  have  presented  their  plans  to  the 
company  with  instructions  to  lay  out  an  indirect  lighting  propo- 
sition. In  numerous  instances  where  a  direct  lighting  system 
has  been  laid  out  by  the  architect,  the  church  authorities  have 
brought  their  plans  to  the  company  to  have  estimates  fur- 
nished for  indirect  lighting,  in  spite  of  the  architect's  drawn 
plans.  There  has  been  practically  no  call  for  semi-indirect  light- 
ing, but  plans  for  semi-indirect  lighting  have  been  drawn  and 
recommended  where,  in  the  opinion  of  the  lighting  specialist,  the 
conditions  were  more  favorable  for  this  method  of  lighting. 

Lighting  Systems. — Direct  lighting  having  been  used  to  the 
greatest  extent  still  predominates.  This  is  largely  due  to  the 
architects,  who  seem  to  be  more  familiar  with  this  method  of 
lighting.     (Fig.  i.) 

There  is  a  tendency,  however,  in  favor  of  indirect  lighting,  or 
a  concealed  direct  lighting  system. 

The  semi-indirect  system  has  not  been  adopted  to  any  extent, 
owing  to  the  greater  cost  of  translucent  glass  bowls  of  large 
dimensions,  and  the  greater  installation  costs. 

Direct  lighting  systems  have  been  installed  more  generally,  due 
to  lower  installation  and  operating  costs,  the  influence  of  the 
architect,  and  the  lack  of  information  pertaining  to  lighting 
matters  on  the  part  of  the  general  public. 

I  do  not  mean  to  favor  any  particular  method  of  lighting; 
the  foregoing  statements  are  based  on  installations  that  have  been 
made  recently. 

It  is  not  at  all  unusual  to  see  new  buildings  and  churches 
equipped  in  a  manner  at  variance  with  good  practice,  simply  be- 
cause the  architects  plans  were  drawn  and  the  client  had  entire 
confidence  in  the  architect.  Generally  architects  will  co-operate 
with  the  illuminating  engineer  but  this  is  not  always  the  case. 


r 


. 


\&yz~-     8*11 1 1 

n  -f     IF  iln  \m  l<Sm  ;- 

9                                    l^2iHfiMm    lis  y»^-Ja 

i  *H   ■  ill  *i  I  I*7"                             1  I      II 

i                         Irai       ]| 

11    IIh.1I  j»SH*!if!! 

wk: 

Bf '■■ 

$^&mb 

GH^'^vTTI 

MM  j    1 

H|     1 

[■Wcaraa}  v*  «o40 

flr  ^ 

-W^^BM— 1— ta—r    l  \   j$k 

u^B                            BP           W-fll  1    a 

|7  *j 

ft  ^ 

Mfjrl; 

ill— -__                       '  "  llA^'A^^'4'^"tff  ■ 

Fig.  5. — Altar  illuminated  with  tungsten -filament  lamps. 


Fig.  6. — An  electric  church  sign. 


5u 

Jo/  June  8f 

(LDRCNS  0/J 

it 

19 

m^k 

ii._t.-.y_- 

1 

Take  P.iit. 

K>:-Z-7? 

COOP 

5/M7/.VC. 

'.  \Su    1 

Fig.  7.— Sign  over  church  entrance. 


Fig.  S. — An  illuminated  bulletin  board. 


Fig.  9. — An  Illuminated  box  sign. 


Fig.  10. — Lighted  entrance  to  church. 


Ely:    church  lighting  617 

General  Requirements  (Auditorium). — There  are  some  gen- 
eral rules  that  should  be  followed,  relative  to  the  illumination  of 
the  church  auditorium. 

No  light  sources  should  be  within  the  line  of  vision. 

All  lamps  in  any  system  of  illumination  should  be  hung  suf- 
ficiently high,  so  they  will  not  obstruct  the  view  of  those  in 
the  gallery  of  the  church. 

No  fixtures  should  be  so  located  that  they  will  detract  from 
the  architectural  features. 

The  intrinsic  brilliancy  of  all  light  sources  visible  from  any 
point  should  be  reduced,  by  frosting  or  other  means. 

Hymn  boards  should  be  illuminated  by  a  concealed  source. 

Chancel. — Generally  in  the  construction  of  the  chancel  one  or 
more  windows  or  sky  lights  are  provided  for  daylight  illumina- 
tion of  the  chancel,  thereby  making  this  section  of  the  church 
lighter  by  comparison  than  the  main  body  of  the  church.  With 
artificial  illumination  this  effect  may  often  be  obtained  more 
readily  and  more  effectively. 

In  formulating  a  plan  or  method  of  lighting  the  chancel,  or 
sanctuary,  the  construction  of  the  interior  of  the  chancel  will  fre- 
quently demand  special  apparatus,  reflectors  and  lamps,  in  order 
to  obtain  effective  results.  A  chancel  arch  will  greatly  facilitate 
the  installation  of  lamps  and  reflectors,  but  when  there  is  no  arch, 
it  is  more  difficult  to  place  the  lamps  and  reflectors  to  direct  the 
light  to  the  best  advantage  and  avoid  sharp  shadows.  Such 
equipment  when  placed  at  a  great  height  should  be  so  arranged 
that  it  can  be  lowered  for  renewal  and  cleaning  purposes.  But 
owing  to  the  installation  costs  lighting  units  are  more  often 
located  in  a  permanent  manner  behind  the  arch  and  renewals  are 
made  by  the  use  of  extension  ladders. 

Altars. — The  use  of  decorative  altar  illumination  is  increasing. 
Electric  imitation  candles  are  being  used  in  many  churches. 
The  lighting  of  tabernacle  niches  as  well  as  the  outlining  of 
altars  is  being  done  quite  successfully  by  the  use  of  low  voltage 
lamps. 

A  sign  lighting  transformer  is  often  installed  and  receptacles 


6l8  TRANSACTIONS   I.    E.    S. — PART    II 

provided  at  various  points  on  the  altar,  together  with  the  perma- 
nent equipment,  so  that  small  electric  bulb  signs  and  emblems 
may  be  readily  connected.  The  cost  of  such  an  installation  is  more 
or  less  offset  by  the  low  operating  costs  of  the  equipment.  For 
instance  in  the  candelabrum  in  Fig.  5  there  were  100  8-candle- 
power  34-watt  carbon  lamps  operating  at  a  cost  of  about  34c. 
an  hour.  They  were  not  used,  except  for  special  services,  on 
account  of  the  cost  of  operation.  Upon  installing  a  sign  lighting 
low  voltage  transformer  and  5-watt  tungsten  lamps  the  cost  was 
reduced  to  about  5c.  per  hour.  This  resulted  in  greater  use  of 
the  equipment.  Of  course  the  illumination  was  reduced,  but  this 
was  offset  by  the  installation  of  two  trough  reflectors  with 
tungsten  lamps  behind  the  sanctuary  columns. 

On  account  of  the  tendency  in  recent  years  to  increase  the  il- 
lumination in  the  body  of  the  church  it  is  necessary  to  increase 
the  illumination  in  the  sanctuary  to  make  a  noticeable  difference 
in  the  intensity  of  the  light  in  the  two  sections  of  the  church. 

The  depth  of  some  of  the  chancels  is  fairly  great  and  in  order 
to  increase  the  illumination  on  the  altar  and  reredos  it  is  necessary 
to  concentrate  the  light  from  a  source  more  or  less  localized. 

Apparatus  and  Fixtures,  etc. — The  use  of  dimmers  in  the 
church  is  becoming  more  general.  In  some  Episcopal  churches 
they  are  placed  on  a  circuit  of  imitation  candles.  As  the  choir 
procession  enters  the  chancel,  the  lamps  are  gradually  brought  to 
full  brilliancy.  In  some  of  the  Catholic  churches  dimmers  are 
connected  to  the  circuits  of  the  church  auditorium  and  sanctuary 
for  use  during  special  services. 

The  question  of  appropriate  fixtures  is  an  important  one. 
The  illuminating  engineer  should  have  a  knowledge  of  decoration 
and  ornamentation,  for  he  is  often  called  upon  to  not  only  plan 
the  outlets  for  a  lighting  scheme,  but  to  select  the  fixtures.  In 
the  case  of  some  of  the  more  novel  schemes  of  illumination,  he 
is  obliged  to  design  fixtures  to  meet  unusual  conditions.  This 
means  that  he  must  be  familiar  with  the  different  orders  of 
architectural  design.  The  illuminating  engineer  should  study 
harmony  of  design,  and  particularly  the  effect  of  light  and  shadow 
on  architectural  ornaments. 

There  have  been  many  apparent  reasons  given  for  copying 


Ely:    church  lighting  619 

ancient  lamps  and  light  sources  to  carry  out  certain  styles  of 
architecture.  In  carrying  out  such  designs  of  fixture  and  lighting 
units,  the  units  have  frequently  been  so  exposed  and  located 
in  such  a  manner  as  to  be  distressing  if  not  harmful  to  the  eyes. 
Therefore,  novel  and  unusual  methods  may  be  inaugurated  in 
designing  the  illumination  for  the  church  with  due  regard  for 
ritualistic  symbolism. 

When  the  illuminating  engineer  attempts  to  effect  new  and 
unusual  methods  of  lighting  churches  he  should  of  course  be 
familiar  with  the  orders  of  architecture,  and  ornament,  so  that 
he  will  be  able  to  so  design  the  lighting  installation  that  some 
of  the  architectural  features  of  the  building  may  be  brought  out 
by  light  and  shadow  without  distorting  their  appearance. 

The  candle  is  used  in  the  services  of  the  Roman  Catholic, 
Episcopal  (high),  Lutheran  and  German  Lutheran  churches. 
It  is  given  more  prominence  in  the  ceremonies  of  the  Roman 
Catholic  church  than  in  the  others. 

In  the  Catholic  churches  the  altar  is  the  main  feature  to  be 
considered.  A  rubric  of  the  church  requires  the  use  of  wax 
candles  for  various  services.  However,  in  numerous  instances 
some  of  the  candles  that  are  still  in  use  can  be  replaced  by  either 
gas  or  electric  imitation  candles  to  greater  advantage,  and  this 
will  result  in  a  more  effective  appearance. 

The  lighting  specialist  should  know  when  and  how  candles 
are  used  in  religious  services,  their  advantages  and  disadvantages. 
The  mellow  glow  of  the  candle  with  the  low  intrinsic  brilliancy 
of  its  flame  is  pleasing,  aside  from  the  flicker.  Whether  the 
flickering  of  the  candles  for  religious  purposes  is  a  distracting 
feature,  or  whether  the  candles  are  more  effective  under  such 
condition  is  probably  an  open  question.  There  are  certain  times, 
particularly  on  holy  and  feast  days,  when  the  altar  is  illuminated 
to  a  greater  extent  than  for  ordinary  services. 

The  candle  has  some  disadvantages  that  cannot  be  overlooked, 
such  as  flicker,  fire  hazard,  and  labor  required  to  maintain  the 
proper  appearance  of  the  candle  and  fixture.  Therefore,  all 
candles  that  are  used  in  addition  to  those  required  by  the  rubrics 
of  the  church  may  be  replaced  with  electric  imitation  candles,  or 
gas  burners. 


620  TRANSACTIONS    I.    E.    S. — PART    II 

Organs. — The  lighting  of  the  keyboard  and  pedals  and  interior 
of  the  organ  frequently  present  some  difficulty,  owing  to  the 
location  of  the  keyboard  and  the  necessity  for  concealing  the 
light  source  from  the  eyes  of  the  congregation.  Some  lighting 
installations  have  been  marred  by  the  glare  of  an  exposed  light 
source  supplying  the  illumination  for  the  music  rack  and  key- 
board, and  by  the  lamps  that  are  used  to  illuminate  the  choir 
when  it  is  located  in  the  front  of  the  church. 

It  is  customary  to  provide  a  portable  lamp  or  extension  bracket 
for  the  music  rack  connected  to  a  live  circuit  to  permit  the  organ- 
ist to  illuminate  his  music  during  his  rehearsal  on  weekdays 
without  using  a  number  of  lamps. 

Illumination  for  the  choir,  when  it  is  located  in  the  gallery, 
can  be  provided  by  brackets  or  ornamental  standards  in  direct 
lighting  installations. 

A  lamp  equipped  with  a  metal  reflector  located  under  the  key- 
board is  frequently  recommended  for  the  illumination  of  the 
pedals. 

ADVERTISING  AND  EXTERIOR  ILLUMINATION. 

Church  advertising  is  being  indulged  in.  A  number  of  churches 
in  Philadelphia  purchase  half  a  page  of  the  newspapers  on  Sat- 
urday of  each  week,  and  announce  subjects,  relative  to  church 
benefits  and  aims,  that  are  to  the  interest  of  the  community. 
This  advanced  step  towards  business  management  of  the  church 
has  greatly  assisted  in  the  introduction  of  exterior  illumination  of 
the  church  and  the  adoption  of  the  electric  sign.  Numerous 
churches  have  installed  both  illuminated  and  straight  electric 
signs.     Some  of  these  signs  contain  large  numbers  of  lamps. 

Greater  attention  should  be  paid  by  the  manufacturer  of  elec- 
tric signs  to  their  designs.  The  sign  should  be  of  a  design  some- 
what similar  to  the  architecture,  and  may  be  improved  by  the  use 
of  Gothic  or  old  English  letters.  The  use  of  these  letters  in 
themselves  would  give  to  the  sign  a  certain  dignity  that  is  not 
obtained  with  the  block  letters  that  are  in  general  use.  But  the 
legibility  of  a  sign  might  thus  be  somewhat  sacrificed.  However, 
this  is  a  detail  which  should  receive  more  favorable  consideration 
from  those  interested  in  the  introduction  of  signs. 

The  illuminated  cross  on  the  steeple  of  the  church  is' very 


ELY  :     CHURCH  LIGHTING 


621 


effective,  and  can  be  seen  at  a  great  distance.     There  is  no  doubt 

that  it  is  an  advertiser  of  the  church. 

Mr.  A.  Larney  describes  in  an  article  on  the  erection  of  an 

illuminated   cross,    a   very    interesting   method    of    installing   the 

lamps  : 

Lamps  are  screwed  into  the  receptacles  mounted  on  three 
flexihle  belts  equipped  with  a  series  of  rope  guides  and  pul- 
leys, by  means  of  which  the  installation  can  he  lowered  from 
the  cross  down  into  the  interior  of  the  spire.  In  this  way 
lamp  renewals  can  he  made  without  the  assistance  of  a 
steeple-jack  or  building  a  scaffold. 


Fig.  11.— Church  exterior  showing  lighting  standards. 


In  spite  of  the  many  drawbacks  and  difficulties  involved  in 
erection  and  maintenance  of  such  crosses,  a  great  many  of  them 
have  been  installed.  The  renewal  of  lamps  lias,  in  numerous 
instances,  been  the  work  of  a  steeple-jack,  but  the  construction, 
as  described  by   Mr.   Larney,  may  be  possible  in  numerous   in- 


622  TRANSACTIONS    I.    E.    S. — PART    II 

stances,  and  should  lead  to  greater  use  of  this  method  of  adver- 
tising the  church. 

The  use  of  lamp  posts  to  improve  the  illumination  of  side- 
walks about  the  exterior  of  the  building  tends  to  make  the 
churches  stand  out  to  better  advantage,  and  to  encourage  attend- 
ance, and  adds  to  the  safety  of  the  pedestrians  when  the  side- 
walks are  covered  with  snow  or  ice. 

Bulletin  Boards. — Some  churches  have  bulletin  boards  that 
may  be  illuminated  at  night.  Special  features  of  this  kind  add 
to  the  value  of  church  lighting.  A  bulletin  board  of  this  kind 
can  be  illuminated  every  night  at  very  small  expense  to  the 
church. 

The  illumination  of  clock-dials  is  a  feature  that  should  not  be 
overlooked,  for  it  adds  to  the  nightly  use  of  light  by  the  church, 
and  acts  as  a  constant  reminder  to  those  traveling  in  the  vicinity. 

Entrance  to  Church. — Entrance  lighting  is  important.  It  not 
only  illuminates  the  stairway  and  the  approach  to  the  church, 
but  indicates  to  the  general  public  when  illuminated  that  services 
are  being  held.  Ornamental  brackets  containing  lamps  are  used 
extensively  for  such  lighting.  Often  the  arch  of  the  entrance  is 
outlined  with  a  series  of  lamps. 

The  importance  of  proper  church  lighting  is  too  often  min- 
imized by  pastors,  church  boards,  and  those  who  have  to  do  with 
church  management.  Ordinary  business  methods  are  about  as 
important  in  the  management  of  the  church  as  they  are  in  a 
mercantile  establishment ;  and  perhaps  more  important  because 
the  objects  to  be  obtained  in  the  case  of  the  church  are  so  much 
more  vital  to  the  entire  community.  It  is  the  business  of  those 
who  manage  church  property  to  see  to  it  that  attendance  at 
church  services  shall  reach  a  maximum,  and  remain  there ;  and 
yet  it  is  surprising  how  little  is  done  to  make  our  churches 
inviting,  cheerful  and  comfortable.  There  is  no  single  factor, 
which  will  accomplish  this  result  so  surely,  and  at  such  small 
initial  expense  as  a  scientifically  designed  system  of  lighting. 

Every  one  knows  that  any  illumination  which  produces  eye- 
strain or  fatigue  among  the  congregation  is  a  positive  force  tend- 
ing to  reduce  the  interest  of  those  attending  church  services.  Pro- 
prietors of  places  of  amusement  discovered  this  fact  many  years 


ELY:     CHURCH    LIGHTING  623 

ago,  and  now-a-days  they  look  upon  the  use  of  light  as  a  posi- 
tive necessity. 

From  every  standpoint  good  illumination  is  just  as  important 
in  the  church  as  it  is  in  the  place  of  amusement.  Emphasis  is 
laid  upon  the  comparative  methods  employed  by  places  of  amuse- 
ment and  churches  in  the  matter  of  making  their  audiences  com- 
fortable. There  is  nothing  in  religion  which  teaches  that 
those  who  attend  church  services  should  be  made  uncomfortable, 
and  yet  lack  of  forethought  upon  the  part  of  those  responsible 
too  often  makes  attendance  at  church  services  a  positive  source 
of  distress  for  those  whose  eyes  are  not  strong  or  cannot  stand 
the  eye-strain  caused  by  poor  illumination. 

DISCUSSION. 

Mr.  J.  R.  Cravat h  :  I  think  it  is  very  gratifying  that  one 
member  here  in  one  year  is  able  to  present  so  many  comparatively 
excellent  examples  of  church  illumination.  I  wish  to  call  your 
attention  to  some  of  the  future  possibilities  of  cove  lighting  with 
the  new  helical  filament  tubular  lamps.  The  cove  lighting  shown 
by  Mr.  Ely  shows  some  undesirable  characteristics  of  cove  light- 
ing as  it  has  necessarily  been  carried  out  in  the  past;  that  is  an 
excessive  amount  of  light  close  around  the  curve  with  too  little 
farther  away.  The  new  tubular  lamp  with  the  helical  filament 
is  going  to  make  it  possible  to  control  the  light  much  better  for 
cove  lighting  and  without  some  of  those  bad  effects  that  have 
heretofore  been  inherent  in  it. 

Fig.  1,  shows  a  church  with  paintings  on  the  ceiling.  I  am 
not  prepared  to  say  that  there  is  any  better  way  of  lighting  that 
particular  church,  but  I  simply  want  to  call  your  attention  to 
the  fact  that  any  exposed  lights  are  going  to  hide  the  paintings 
to  a  certain  extent,  and  that  this  fact  must  be  borne  in  mind  by 
the  architect  when  he  is  designing  the  church.  If  it  is  so  de- 
signed that  it  must  necessarily  be  lighted  by  some  exposed  light- 
ing unit,  the  effect  of  the  paintings  will  be  lost  at  night. 

Mr.  A.  L.  Powell:  Church  lighting  forms  an  extremely 
interesting  subject  and  an  observation  of  most  of  the  existing 
installations  shows  that  there  is  still  a  great  deal  of  work  to  be 
done. 


624  TRANSACTIONS   I.    E.    S. — PART   II 

By  way  of  supplementing  the  data  included  in  the  paper,  the 
following  remarks  may  be  of  value. 

As  a  result  of  a  number  of  tests,  it  has  been  found  that  an 
average  intensity  of  1.5  foot-candles  on  a  3-foot  horizontal  plane, 
is  very  satisfactory  for  church  lighting,  where  proper  precau- 
tions have  been  taken  to  shield  the  eyes  from  glare. 

Semi-indirect  illumination  is  quite  feasible  and  the  cost  has 
not  been  high.  There  are  simple  designs  in  pressed  opalescent 
glass  bowls  whch  give  excellent  results,  and  the  cost  is  very 
slight.  The  use  of  white  leaded  glass  semi-indirect  units  is  a 
very  promising  innovation,  and  it  is  possible  to  design  units  along 
Gothic  or  Renaissance  lines,  so  that  they  will  harmonize  ex- 
cellently with  the  church  architecture.  I  have  in  mind  a  church 
in  North  Adams,  Mass.,  where  there  are  large  fixtures  of  white 
leaded  glass  of  Gothic  design  6  feet  in  diameter  which  hang 
15  feet  from  the  peak  of  the  arch.  Each  fixture  has  3  150-watt 
and  3  250-watt  tungsten  filament  lamps.  Only  a  few  built  up 
units  of  this  type  are  necessary  for  a  given  space  and  conse- 
quently the  wiring  cost  may  be  reduced  considerably. 

This  white  glass  has  been  utilized  for  direct  lighting.  An 
asymmetrical  reflector  used  in  the  Buffalo  General  Electric  Com- 
pany's building,  was  described*  by  Mr.  Ryan,  at  last  year's  con- 
vention. The  design  was  slightly  modified  and  the  equipment 
used  for  a  church  which  had  a  fan-shaped  roof  that  reached 
a  maximum  height  above  the  pulpit.  Outlets  were  located  on  the 
ribs  and  the  fixtures  hung  to  direct  the  maximum  light  toward  the 
front.  There  was  sufficient  diffused  light  transmitted  through 
the  glass  to  light  the  balcony,  and  the  main  portion  was  lighted 
by  the  reflected  light. 

The  term  "line  of  vision"  is  mentioned  in  the  paper,  and  that 
brings  to  mind  the  question:  What  is  the  line  of  vision?  As  a 
homely  illustration,  one  may  note  that  the  ladies  of  the  church 
often  wonder  why  the  gentlemen  go  to  sleep  so  much  more 
readily  than  they  do  at  a  service.  It  has  been  my  experience 
that  there  is  usually,  within  the  angles  of  vision,  a  number  of 
bare  or  improperly  shielded  light  sources,  which  shine  directly 
into  my  eyes.     The  ladies  have  their  hats  on  while  in  church, 

*  Trans.  I.  E.  S.,  Vol.  VII,  No.  8  (Nov.,  1912),  p.  597. 


CHURCH    LIGHTING  625 

and  hence,  their  eyes  are  more  completely  protected,  and  drowsi- 
ness does  not  result  as  readily. 

As  regards  the  designing  of  special  reflectors  for  chancel 
lighting,  I  may  say  that  we  have  had  very  good  suc- 
cess in  lighting  chancels  of  the  arch  type,  by  using  angle 
steel  or  glass  reflectors,  giving  asymmetric  curves.  In  the  chan- 
cel without  the  arch,  there  are  quite  often  Corinthian  columns 
or  similar  structures  on  both  sides,  and  by  using  the  tubular  type 
lamp  with  a  cylindrical  or  trough  reflector,  to  direct  the  light  to 
the  opposite  wall  rather  than  to  the  adjacent  wall,  as  in  cove  light- 
ing, good  results  have  been  obtained.  When  lighted  from  the 
side  rather  than  from  overhead  most  chancels  appear  better 
illuminated  because  the  shadow  effects  are  softened. 

The  lighting  in  the  choir  loft  is  often  very  annoying,  for  it  is 
usually  accomplished  by  a  number  of  small  lamps  with  diffusing 
shades  which  are  in  the  line  of  vision.  In  some  churches  where 
it  has  been  impossible  to  install  lamps  within  the  interior  of  the 
organ,  or  on  the  music  racks,  the  opaque,  bowl  steel  reflector, 
with  its  exterior  painted  to  harmonize  with  the  woodwork,  has 
been  used. 

Window  lighting  from  the  exterior  has  been  very  satisfactorily 
accomplished  by  the  use  of  asymmetrical,  weatherproof  type  steel 
reflectors  and  regular  lamps,  or  the  tubular  type  lamp  and  small 
reflectors  arranged  to  evenly  illuminate  the  entire  window  surface. 

Prof.  F.  C.  CaldweXl  :  The  point  was  brought  out  that 
the  deck  lighting  as  shown  on  Fig.  2,  does  not  give  suitable  light- 
ing for  the  upper  part  of  the  room.  This  is  perhaps  due  to  the 
fact  that  the  deck  is  somewhat  recessed — the  glass  should  be  kept 
down  as  near  the  ceiling  as  possible.  In  deck  lighting  it  is  impor- 
tant that  a  glass  of  good  transmission  efficiency  be  used.  Ordi- 
nary skylight  glass  is  very  well  for  daylight,  but  seriously  inter- 
feres with  the  efficiency  of  a  lighting  system.  Even  where  econ- 
omy is  not  of  prime  importance  the  best  results  can  usually  be 
obtained  by  putting  money  into  good  glass  rather  than  into 
additional  power. 

Mr.  R.  F.  Pierce  :  There  were  one  or  two  points  brought  out 
in  these  two  papers  that  are  of  especial  interest  in  connection 
with  the  paper  presented  yesterday  by  Dr.  Ferree.    There  seems 


626  TRANSACTIONS    I.    E.    S. — PART    II 

to  be  a  considerable  movement  in  the  direction  of  reducing  the 
brightness  of  surfaces  within  the  range  of  vision.  This  is  shown 
by  the  popularity  of  direct  and  semi-indirect  systems  of  lighting. 
In  churches,  particularly,  people  are  quite  sensitive  to  the  aggra- 
vation produced  by  glaring  light  sources,  and  the  fact  that  they 
have  often  resorted  to  the  so-called  indirect  system  of  lighting 
is  excellent  evidence  that  our  commercial  glassware  which,  the 
reflector  manufacturer  tells  us  effectively  shades  the  lamp  and 
reduces  the  glare,  really  does  comparatively  little  to  that  end. 
The  results  given  in  Dr.  Ferree's  paper  yesterday  indicate  that 
reducing  surface  brightness  of  reflectors  to  the  extent  commonly 
found  in  commercial  types  is  not  sufficient  to  make  the  surface 
much  less  objectionable  from  a  standpoint  of  depreciation  in  eye 
efficiency  than  the  bare  lamp.  We  would,  however,  not  be  justi- 
fied in  concluding  that  indirect  lighting  as  such  is  responsible  for 
the  results  obtained  by  Dr.  Ferree.  If  we  use  a  direct  lighting 
system  in  which  the  enclosing  glassware  forms  the  whole  ceiling 
of  the  room  as  shown  in  Fig.  2  in  Mr.  Ely's  paper  and  in  the 
installation  described  by  Mr.  Kingsbury,  we  have  an  installation 
in  which  the  surface  brightness  of  the  glassware  is  no  higher 
than  that  of  the  ceiling  in  an  indirect  lighting  system  to  produce 
the  same  illumination,  and  we  would  certainly  not  expect  any 
different  results  as  regards  the  efficiency  of  the  eye.  On  the 
other  hand,  if  we  use  an  indirect  reflector  concentrating  the  light 
on  a  very  small  spot  on  the  ceiling  and  producing  a  surface 
brightness  in  the  neighborhood  of  two  or  three  candle-power 
per  square  inch,  we  would  have  a  condition  that  we  would  expect 
to  be  quite  as  annoying  and  quite  as  unfavorable  to  the  efficiency 
of  the  eye  as  a  direct  lighting  system  in  which  a  similar  distri- 
bution of  surface  brightness  occurred.  It  appears  that  the  prin- 
cipal factor  is  the  brightness  and  area  of  the  illuminated  sur- 
faces, and  it  is  immaterial  whether  a  certain  distribution  is 
obtained  by  "direct"  or  "indirect"  means. 

One  serious  drawback  to  the  indirect  system  is  the  reversal 
of  the  natural  order  of  intensities.  Under  daylight,  the  higher 
intensities  are  found  at  the  lower  levels,  and  the  lower  intensi- 
ties at  the  upper  levels.  This  condition  is  reversed  with  indirect 
lighting  on  account  of  the  fact  that  the  reflecting  surfaces  are 


CHURCH    LIGHTING  627 

generally  diffusing,  each  element  giving  a  circular  distribution 
curve,  and  it  is  practically  impossible  to  avoid  illuminating  the 
upper  portions  of  the  side  walls  to  a  higher  degree  than  the  lower 
portions. 

This  effect  may  be  avoided  in  deck  lighting  systems,  however, 
since  the  deck  may  be  constructed  of  glass  which,  while  suffi- 
ciently diffusing  to  reduce  surface  brightness,  will  not  seriously 
interfere  with  the  direction  given  to  the  light  rays  by  the  reflec- 
tors. This  effect  is  seen  in  the  illustration  of  the  deck  lighting 
systems  shown  by  Mr.  Ely. 

Another  important  consideration  in  church  lighting  which  fre- 
quently militates  against  the  employment  of  indirect  lighting  is 
that  of  esthetic  effect.  The  prevailing  type  of  church  architec- 
ture is  Gothic,  in  which  it  is  the  purpose  of  the  architect  to  allow 
the  high,  pointed  arches  to  remain  in  comparative  darkness. 
When  the  indirect  system  of  lighting  is  used,  the  brilliant  illumi- 
nation in  this  portion  of  the  building  entirely  destroys  the  effect 
which  the  architect  strived  to  produce.  Some  particularly  atro- 
cious examples  of  a  disregard  for  architectural  considerations  by 
the  application  of  indirect  lighting  to  Gothic  interiors  have  been 
found  in  recent  installations,  and  are  excellent  examples  of  what 
good  illumination  should  not  be. 

In  the  installation  described  by  Mr.  Kingsbury,  use  was  made 
of  what  might  be  termed  a  semi-indirect  lighting  system  in 
which  the  wall  is  used  as  a  reflecting  surface  instead  of  the  ceil- 
ing. This  is  quite  similar  to  one  of  the  installations  described 
by  Dr.  Ives  in  reporting  his  experiments  on  the  approximation 
of  daylight  distribution  in  residence  interiors.  As  I  have  not 
had  an  opportunity  to  observe  the  results  of  lighting  of  this 
character,  I  am  unable  to  comment  upon  it ;  but  I  think  it  presents 
a  problem  worthy  of  more  extended  investigation. 

Mr.  J.  R.  Cravat  11 :  In  regard  to  Fig.  2  in  Mr.  Ely's  paper, 
I  don't  agree  with  Mr.  Pierce  that  it  represents  an  ideal  condi- 
tion, because  just  as  Mr.  Luckiesh  has  said,  the  contrast  of  the 
brightness  within  the  range  of  vision  is  the  very  important  point. 
In  this  case  if  we  are  to  judge  from  the  photograph  (which, 
however,  may  not  represent  things  just  as  they  are),  we  have 


628  TRANSACTIONS   I.    E.    S. — PART   II 

a  very  decided  contrast  between  the  brightness  of  the  skylight 
and  the  brightness  of  the  ceiling  and  bright  surroundings.  Now, 
that  is  probably  about  the  best  way  that  particular  installation 
could  be  lighted,  but  it  does  not  illustrate  an  ideal  condition, 
because,  that  contrast  must  be  annoying.  Any  kind  of  violent 
contrast  of  surface  brightness  which  one  must  face  constantly 
cannot  fail  to  be  annoying. 

Mr.  T.  J.  LiteE,  Jr.  :  In  a  number  of  the  installations  de- 
scribed by  Mr.  Ely,  particularly  the  one  shown  in  Fig.  2,  it 
would  appear  that  mural  decorations  on  the  side-walls  and  ceil- 
ings are  not  properly  lighted.  The  decorations  referred  to 
are  usually  very  costly  and  in  the  lighting  of  such  a  building 
the  illumination  must  not  only  be  sufficient  for  reading  purposes 
but  must  be  of  a  character  which  will  properly  bring  out  the 
decoration  above  referred  to. 

In  reference  to  Fig.  1,  I  should  say  that  a  person  sitting  half- 
way back  in  the  church  would  see  at  least  half  of  the  fixtures. 
In  other  words,  the  light  sources  would  be  within  his  range  of 
vision.  I  think  the  lighting  of  a  church  should  be  so  arranged  that 
the  lamps  would  be  out  of  the  range  of  sight.  The  light  would 
probably  be  in  the  speaker's  eyes,  but  this  could  not  be  avoided. 
At  any  rate,  he  is  more  apt  to  look  down  upon  the  congregation 
and  with  the  lamps  hung  high  in  the  church  they  would  not  be 
so  apt  to  annoy  him,  and  even  if  they  did,  from  the  standpoint 
of  the  greatest  good  to  the  greatest  number  such  an  arrangement 
would  be  considered  preferable. 

Why  should  the  lighting  of  a  church  building  be  considered 
so  differently  from  the  lighting  of  a  theatre?  In  the  latter  case, 
the  lamps  are  of  necessity  shining  in  the  player's  eyes.  Bare 
lamps  are  never  allowed  to  annoy  the  audience. 

Mr.  E.  B.  RowE:  I  have  one  or  two  questions  to  ask.  One 
Mr.  Powell  has  touched  on  in  connection  with  the  use  of  semi- 
indirect  lighting.  On  the  sixth  page  Mr.  Ely  mentions  the  fact 
that  semi-indirect  lighting  has  been  recommended  where  the  condi- 
tions were  deemed  very  favorable.  I  would  like  as  a  matter  of 
information  to  have  him  give,  if  he  can,  what  he  considers  the 
conditions  which  are  favorable  to  that  system  more  than  condi- 
tions which  are  not. 


CHURCH    LIGHTING  629 

In  regard  to  the  use  of  light  units  on  the  chancel  side  of 
beams,  etc.,  it  occurred  to  me  that  there  is  one  disadvantage  in- 
herent in  that  type  of  lighting  and  that  is  the  glare  in  the  eyes 
of  the  pastor  and  the  choir  if  it  is  on  the  chancel  end  of  the 
nave;  perhaps  he  may  have  some  information  to  give  us  as 
to  whether  there  have  been  any  objections  made  from  that 
standpoint.  That  is  an  excellent  method  I  think  of  obtaining  an 
efficient  illumination  very  similar  to  direct  lighting. 

It  occurred  to  me  with  regard  to  the  lighting  of  croses  on  the 
exterior  of  the  church,  which  sometimes  have  to  be  located  in 
inaccessible  points,  that  use  might  be  made  of  the  new  con- 
centrated filament  incandescent  lamp  in  parabolic  reflectors,  so 
that  there  would  be  no  need  of  ever  going  up  to  the  cross  itself. 

Mr.  LuckiESH  :  I  want  to  supplement  one  of  the  points 
brought  out  by  Mr.  Pierce  by  describing  one  of  the  most  annoy- 
ing cases  of  glare  I  ever  experienced.  This  installation  is  in  one 
of  the  modern  auditoriums  in  Cleveland  which  was  installed  by 
an  architect  who  I  know  has  little  use  for  an  illuminating 
engineer.  The  auditorium  proper  is  lighted  from  beautiful  in- 
direct fixtures.  The  pulpit  however  is  lighted  by  concealed 
sources  as  in  Fig.  2,  of  Mr.  Ely's  paper.  When  the  pulpit  alone 
is  lighted  the  contrast  between  its  bright  background  and  the 
dark  surroundings  causes  a  most  annoying  glare.  This  brings 
out  the  point  that  glare  is  not  always  due  to  high  brightness. 

Mr.  G.  H.  Stickney:  The  author  is  to  be  congratulated  on 
having  such  a  large  number  of  successful  church  lighting  installa- 
tions. There  is  probably  no  other  class  of  lighting  problem 
which  the  illuminating  engineer  approaches  with  more  trepida- 
tion, since  the  artistic  requirement  in  church  lighting  is  so  pre- 
dominant that  the  engineer,  unless  he  can  fully  co-operate  with 
the  architect,  is  at  a  tremendous  disadvantage. 

One  of  the  most  interesting  problems  in  church  lighting  which 
I  have  ever  handled,  and  one  which  illustrates  a  novel  method 
which  I  have  not  seen  used  elsewhere,  was  in  a  large  Gothic 
cathedral  in  Montreal.  This  is  a  magnificent  building  erected 
about  100  years  ago.  The  main  portion  consists  of  three  Gothic 
naves,  the  center  one  being  about  100  feet  high.  As  the  ceiling 
is  dark  finish,  the  usual  indirect  lighting  would  have  been  unsuit- 


63O  TRANSACTIONS   I.    E.    S. — PART   II 

able,  although  the  equipment  actually  used  might  be  classed  as 
indirect.  In  the  previous  installation  fixtures  were  suported 
from  the  backs  of  the  seats  and  fairly  brilliant  light  sources 
located  about  7  or  8  feet  above  the  floor  in  such  a  manner  as  to 
detract  from  the  general  view,  especially  of  worshippers  in  the 
rear  seats.  In  the  upper  part  of  the  building  along  each  side 
of  the  central  nave  was  a  line  of  windows  which  opened  into  a 
covered  space  between  the  nave  and  the  roof.  Prismatic  glass 
was  inserted  in  these  windows  and  large  tungsten  filament  lamps 
arranged  behind  them,  each  equipped  with  a  metal  reflector  so 
as  to  direct  the  light  through  the  window.  The  prism  glass 
deflected  the  light  downward,  and  it  was  possible,  by  adjusting 
the  height  of  the  lamps  with  regard  to  the  windows,  to  control 
the  distribution  of  light  and  proportion  it  properly  between  the 
upper  and  lower  portions  of  the  church.  The  energy  consump- 
tion was  a  little  over  one  watt  per  square  foot  of  floor  area. 
While  I  never  had  an  opportunity  of  seeing  the  completed  instal- 
lation, it  has  been  reported  as  producing  a  most  pleasing  effect, 
and  that  it  is  possible  to  read  in  any  part  of  the  church. 

Mr.  R.  B.  Ely  :  In  answer  to  Mr.  Cravath's  remark  about 
the  installation  in  which  the  paintings  appear — In  that  case  we 
used  a  cluster  of  tungsten  lamps  equipped  with  distributing  type 
reflectors  of  light  opal  glass  to  get  diffusion,  which  was  attached 
to  a  rope  and  pulled  up  from  the  floor  to  the  ceiling;  it  was 
placed  at  various  heights  until  we  got  that  height  at  which  the 
paintings  could  be  seen  to  best  advantage.  By  increasing  the 
intensity  (various  sizes  of  lamps  were  tried)  so  as  to  get  more 
brilliancy  all  over  the  church,  it  was  found  that  the  paintings 
could  be  seen  more  readily  with  a  higher  intensity  than  under 
the  lower  intensity. 

Mr.  Powell  touched  on  the  intensity  of  illumination.  It  should 
range,  he  stated,  from  ^4  to  1.5  foot-candles.  However,  some 
churches,  particularly  Catholic  churches,  have  to  be  lighted  as 
brilliantly  as  possible  at  certain  times,  Christmas  and  Easter  for 
example.  Sufficient  equipment  should  be  installed  to  provide  the 
extra  illumination  required  on  such  occasions. 

Mr.  Powell  spoke  about  the  question  of  lamps  being  in  the 


CHURCH    LIGHTING  63 1 

line  of  vision,  and  answered  it  himself.     I  think  we  can  all  tell 
when  a  lamp  is  annoying  or  whether  it  is  in  the  line  of  vision. 

Reference  was  made  to  special  reflectors  for  chancel  illumina- 
tion. There  have  been  cases  where  it  has  been  desirable  to 
illuminate  the  chancel  more  brilliantly  than  other  sections  of  the 
church,  to  bring  out  that  part  of  the  church.  And  in  such  cases 
it  is  desirable  to  cut  off  the  illumination  from  the  lower  and  top 
sides  of  fixtures  to  get  that  effect.  We  have  in  some  instances 
made  special  corrugated  glass  reflectors  with  definite  cut-off  points. 

Reference  has  also  been  made  to  putting  lamps  behind  Corin- 
thian columns,  with  tubular  lamps.  We  were  called  in  on  a  case 
that  had  such  equipment,  simply  because  the  desired  results 
could  not  be  obtained.  The  church  panels  which  had  been  in 
stalled  at  a  cost  of  something  like  $2,000  were  entirely  flat  with 
this  method  of  lighting;  no  irridescence  from  the  tile  panel  was 
to  be  had.  Search-light  lamps  with  parabolic  reflectors  were 
used  to  direct  light  on  these  panels  until  the  angle  where  the  ir- 
ridescence would  appear  to  the  congregation  was  found.  Then 
equipments  were  installed  at  those  points,  and  the  lighting  was 
found  to  be  very  effective. 

In  window  lighting,  as  a  rule,  the  main  window  or  the  most 
beautiful  window  frequently  appears  at  the  rear  of  the  church. 
And  the  equipment  for  the  gallery  is  generally  located  in  that  sec- 
tion. With  indirect  and  semi-indirect  systems  the  fixture  for  this 
portion  of  the  church  is  usually  located  so  that  it  will  illuminate 
that  window  to  the  best  advantage,  and  at  the  same  time  provide 
illumination  for  the  gallery. 

Regarding  Mr.  Caldwell's  question  about  the  introduction  of 
semi-indirect  lighting — there  are  quite  a  number  of  cheap  bowls 
for  semi-indirect  light,  but  when  one  is  dealing  with  a  church 
where  generally  very  large  units  have  to  be  used,  I  have  invari- 
ably found  the  cost  to  be  greater  than  that  of  other  types.  It 
doesn't  make  so  much  difference  if  it  is  a  new  installation,  that 
is  in  a  new  building  being  erected ;  'the  cost  is  then  not  such  a 
factor.  But  if  it  is  an  old  installation  and  there  is  competition 
it  is  a  factor. 

Mr.  Pierce  referred  to  Fig.  2.    The  effect  is  not  as  it  appears 
in  the  photograph.     There  is  considerable   diffusion   from   the 
6 


632  TRANSACTIONS   I.    E-    S. — PART   II 

walls  which  are  very  light  buff;  and  the  paneling  of  the  ceiling 
may  be  seen  readily.  It  is  not  dark,  as  it  appears  in  this  photo- 
graph; the  darkness  is  probably  due  to  long  exposure  in  taking 
the  picture. 

Mr.  Luckiesh  referred  to  the  glare  from  the  chancel.  When 
the  rest  of  the  lights  in  the  auditorium  are  turned  out  we  try 
as  far  as  possible  to  get  a  theatrical  appearance,  you  might  say; 
that  is,  to  concentrate  the  light  on  the  pastor,  as  you  would  in 
a  theater  concentrate  the  light  on  the  actors.  And  frequently 
where  there  is  a  very  light  background  a  portion  of  the  lights 
are  turned  out,  excepting  those  that  would  tend  to  show  the 
pastor. 

Mr.  Litle  commented  on  the  same  effect  as  produced  in  Fig.  2, 
and  as  to  reflection  from  walls,  ceilings  and  decorations.  These 
are  all  well  brought  out  under  that  installation.  When  planning 
a  lighting  system  for  a  church  we  consider  the  decorations.  They 
constitute  a  feature  which  should  be  properly  illuminated. 

I  believe,  as  Mr.  Marks  said  yesterday,  that  "any  system  of 
illumination  can  be  made  very  effective,"  and  in  all  church  instal- 
lations we  are  largely  governed  by  the  architectural  considera- 
tions and  the  character  of  walls  and  ceilings. 


EDWARDS   AND    HARRISON  I    ACCURACY    OF    PHOTOMETRY      633 

SOME  STUDIES  IN  ACCURACY  OF  PHOTOMETRY.* 


BY    EVAN    J.    EDWARDS    AND    WARD    HARRISON. 


Synopsis:  Five  separate  investigations  are  reported  in  this  paper: 
I — Relative  accuracy  of  Bunsen  and  Lummer-Brodhun  devices.  This  test 
involving  several  thousand  readings  showed  the  sensitivity,  expressed  in 
average  deviation,  to  be  0.4  per  cent,  for  Lummer-Brodhun  as  compared 
with  1.5  per  cent,  for  the  Bunsen.  II — Magnitude  of  error  due  to  parallel- 
ism of  rays,  in  the  photometry  of  reflector  sources.  The  results  of  tests 
on  typical  reflectors  for  general  illumination  purposes  show  the  errors  to 
be  negligible.  Ill — Method  of  investigating  adjustment  error;  and  the 
calibration  of  portable  photometers.  This  method  consists  in  taking  pho- 
tometer readings  for  various  distances  from  a  working  standard  lamp  and 
analyzing  the  results  by  reducing  the  relation  to  a  straight  line  function. 
The  constant  of  the  photometer  is  obtained  from  the  slope,  and  working 
standard  intensity.  IV — Errors  in  illumination  measurements  due  to  fail- 
ure of  test  plate  to  obey  the  cosine  law.  Discrepancies  between  measured 
and  calculated  values  of  illumination  are  fully  accounted  for  by  this  inves- 
tigation. The  average  error  of  the  plates  investigated  was  found  to  be 
over  10  per  cent,  at  450.  Computations  show  that  the  photometer  results 
for  an  average  installation  are  about  12  per  cent.  low.  V — Method  of 
obtaining  and  recording  distribution  data.  A  so-called  thousand  lumen 
basis  of  drawing  distribution  curves  is  proposed,  in  order  to  avoid  error 
and  confusion  in  comparing  reflector  units.  On  this  basis  zonal  lumen 
values  are  instantly  convertible  to  per  cent,  of  total  and  intensity  values 
to  per  cent,  of  horizontal  for  the  bare  lamp.  By  using  a  single  multiply- 
ing factor  all  values  can  be  corrected  to  current  lamp  efficiencies. 

Many  investigations  which  have  to  do  with  accuracy  in  photo- 
metry have  been  made  in  the  laboratories  with  which  the  authors 
of  this  paper  are  identified.  These  investigations  have  served 
their  purpose  as  far  as  the  particular  laboratory  is  concerned, 
but  in  some  instances  have  not  been  reported.  It  is  believed 
that  brief  reviews  of  the  more  important  ones  will  prove  useful 
to  the  members  of  the  Illuminating  Engineering  Society. 

This  paper,  then,  is  of  the  nature  of  a  report  on  five  separate 
investigations. 

*  A  paper  read  at  the  seventh  annual  convention  of  the  Illuminating  Engineering 
Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 

The  Illuminating  Engineering  Society  is  not  responsible  for  the  statements  or 
opinions  advanced  by  contributors. 


634  TRANSACTIONS   I.    E-    S. — PART    II 

I — Relative  accuracy  of  Bunsen  and  Lummer-Brodhun  de- 
vices. 

II — Magnitude  of  error  due  to  parallelism  of  rays,  in  the 
photometry  of  reflector  sources. 

Ill — Method  of  investigating  adjustment  error;  and  the  cali- 
bration of  portable  photometers. 

IV — Errors  in  illumination  measurements  due  to  failure  of 
test  plate  to  obey  the  cosine  law. 

V. — Method  of  obtaining  and  recording  distribution  data. 

No  attempt  is  made  to  connect  the  various  investigations,  al- 
though it  will  be  seen  that  the  method  used  in  Investigation  2, 
suggested  the  procedure  of  Investigation  3. 

I.   RELATIVE  ACCURACY  OF  BUNSEN  AND  LUMMER- 
BRODHUN  DEVICES. 

There  is  at  present  little  question  as  to  the  order  of  sensitivity 
of  the  common  forms  of  photometric  devices.  The  quantitative 
results  of  an  extended  test  to  obtain  relative  accuracy  values 
of  the  Bunsen  and  Lummer-Brodhun  devices,  which  resulted  in 
a  decision  to  discard  all  Bunsen  apparatus  in  favor  of  Lummer- 
Brodhun,  may,  however,  be  of  value. 

About  five  thousand  readings  extending  over  a  period  of 
three  months  were  taken  in  order  to  establish  beyond  question 
the  relative  accuracies  of  the  two  devices  under  the  particular 
conditions  involved.  The  Bunsen  photometer  used  was  of 
the  regular  type,  a  circular  spot  surrounded  by  a  concentric  field 
and  viewed  by  means  of  two  angle  mirrors.  The  Lummer- 
Brodhum  photometer  used  was  of  the  low-contrast  variety,  which 
has  been  shown*  to  be  more  sensitive  than  the  earlier  high  con- 
trast type.  Except  for  the  sight  box,  the  same  photometric 
equipment  was  used  for  the  entire  test.  Twenty  incandescent 
electric  lamps  of  various  kinds  and  efficiencies  were  used,  all 
against  the  same  comparison  lamp  in  order  to  get  an  idea  as  to 
the  effect  of  color  difference. 

The  details  of  the  test  and  many  interesting  but  less  important 
deductions  must  be  omitted.  As  an  example  may  be  cited  the 
test  on  the  effect  of  changing  the  shape  of  the  Bunsen  spot. 
It  was  found  that  a  star  shaped  spot  was  easier  to  manipulate 

*  I^unimer  and  Brodhun,  Zeitschrift  fur  Instrumenltnkunde,  Vol.  9,  p.  461. 


EDWARDS    AND    HARRISON  :    ACCURACY    OF    PHOTOMETRY      635 


and  showed  a  higher  sensibility  although  of  the  same  order. 
It  is  sufficient  here  to  show  the  condensed  results  for  the  more 
common  type  of  Bunsen  in  the  curves  of  Fig.  I  bearing  in  mind 
that  the  precision  is  expressed  in  average  deviation  from  the 
mean  and  that  ten  readings  were  taken  on  each  lamp  with  each 
device  for  each  set  of  observations.  It  is  seen  that  the 
Lummer-Brodhun  precision  is  about  four  times  as  good  as  the 
Bunsen.  The  actual  grand  average  for  the  entire  test  is  0.4  per 
cent,  average  deviation  for  the  Lummer-Brodhun  and   1.5  per 


Fig.  1. — Relative  accuracy  of  Bunsen  and  L,utnmer-Brodhun  devices. 

cent,  for  the  Bunsen.  No  marked  loss  in  precision  results 
from  small  color  differences  such  as  with  carbon  against  tung- 
sten. 

II.  MAGNITUDE  OF  ERROR  DUE  TO  PARALLELISM  OF    RAYS, 
IN  THE  PHOTOMETRY  OF  REFLECTOR  SOURCES. 

Many  who  have  had  occasion  to  calculate  illumination  from 
the  distribution  curves  of  lamps  with  concentrating  reflectors 
have,  no  doubt,  had  a  feeling  of  uncertainty  as  to  the  accuracy 
of  their  results,  due  to  the  possible  parallelism  of  a  portion  of 
the  light  rays. 

It  would  seem  that  where  a  reflector  directs  the  rays  to  a 
considerable  degree,  that  there  would  be  brought  about  an  ap- 
preciable effect  of  parallelism  of  the  rays.  This  investigation 
was  undertaken  with  the  idea  of  obtaining  a  measure  of  this 
effect  of  parallelism  in  commercial  types  of  reflector  units. 

In  the  theoretical  extreme  case  of  a  parabolic  reflector  and  a 
point   source,   the   illumination   at  various   distances    would   be 


636  TRANSACTIONS   I.   E.    S. — PART   II 

constant.  The  point  source  with  no  reflector  would,  of  course 
furnish  an  illumination  which  would  be  strictly  proportional  to 
the  reciprocal  of  the  square  of  the  distance.  A  partial  parallel- 
ism, such  as  results  from  the  use  of  a  directive  reflector,  would 
be  expected  to  bring  an  illumination  which  does  not  decrease 
with  distance  as  much  as  would  be  given  by  the  inverse  square 
law. 

In  order  to  determine  the  magnitude  of  these  effects,  tests 
were  conducted  on  a  number  of  typical  reflectors  using  a  Weber 
portable  photometer  in  such  a  manner  as  to  eliminate  other 
sources  of  error.  The  method  employed  depends  upon  the 
principle  that  when  a  photometer  screen  is  balanced  between 
two  point  sources  of  constant  intensity  the  ratio  of  the  distances 
of  the  two  lamps  from  the  screen  is  constant.  Readings  were 
first  taken  on  a  bare  lamp  at  distances  varying  from  3  to  25  feet 
0.914  to  7.315  m.)  to  serve  as  a  test  of  the  adjustment  of  the 
photometer,  and  then  the  bare  lamp  was  replaced  by  various 
reflector  units  in  turn,  and  readings  obtained.  The  relation 
between  the  test  and  comparison  lamp  distances  from  the  photo- 
metric screens  is  a  straight  line  for  point  sources,  and  moreover 
a  straight  line  passing  through  the  origin.  Therefore  a  plot 
of  distance  of  the  test  unit  against  photometer  reading  (since 
a  Weber  photometer  reads  comparison  lamp  distance  directly) 
serves  as  a  complete  test  of  the  inverse  square  law  is  applied 
to  the  unit  in  question. 

The  distance  of  the  light  unit  was  varied  by  raising  and  lower- 
ing it,  the  portable  photometer  being  set  up  directly  beneath. 
Readings  were  taken  in  the  direction  of  the  axis  of  symmetry  of 
the  unit,  since  the  effect  should  be  most  pronounced  at  this 
angle. 

The  results  for  several  types  of  reflector  units  are  summarized 
in  the  curves  of  Fig.  2.  They  are  self-explanatory,  both  as  re- 
gards quantitative  results  and  their  precision  since  the  actual 
points  are  shown.  Each  point  is  obtained  from  the  average 
of  four  photometer  readings.  The  abscissae  of  the  bare  lamp, 
extensive  and  intensive  curves,  have  been  multiplied  by  two  in 
order  to  show  all  curves  conveniently  on  one  sheet.     Therefore, 


EDWARDS   AND    HARRISON  '.   ACCURACY   OF   PHOTOMETRY      637 

the  intercepts  of  the  extensive  and  intensive  curves,  as  drawn, 
show  twice  their  actual  values. 

It  is  seen  that  the  curves  are  fairly  good  straight  lines  and 
also  that  the  intercept  is  within  the  body  of  the  filament  in 
every  case. 

A  failure  of  the  line  to  pass  through  the  origin  shows  that 
the  distance  which  should  be  used  in  computing  illumination  by 
the  inverse  square  law  may  not  be  exactly  the  same  as  the  dis- 
tance to  the  center  of  gravity  of  the  filament.  It  appears  from 
the  curves  that  such  a  discrepancy  is  more  likely  to  occur  than 


?/ 

- 

/ 

f 

r 

i 

- 

v 

i 

j 

^ 

,-- 

• 

•" 

\ 

I 

A 

1 

y 

-y 

<■- 

- 

* 

>' 

- 

9 

> 

> 

t 

- 

- 

/ 

y 

t 

/ 

• 

- 

-- 

: 

l  =y 

Si 

^ 

i 

.-• 

- 

* 

--' 

31 

S£ 

I    % 

N 

/ 

t 

r 

- 

r 

y 

£ 

9 

tf» 

it 

.- 

* 

J 

•-.'. 

J« 

IM 

i£rv 

?.S 

y 

t 

•* 

*$ 

n 

mz 

m 

oJ 

■ 

/ 

• 

> 

ta 

a 

r-7_ 

.tlis 

'■ 

a 

■  ■ 

::-'. 

-i . 

■    c 

«  1 

- 

'■  - 

"  _ 

;". 

--.   ! 

-V- 

';; 

,.■= 

M 

r 

' 

c,-    j 

Fig.  2. — Curves  of  photometric  readings  and  distances  for  reflectors,  showing  relative 
intercepts.  Data  obtained  with  100-watt  tungsten  lamp.  (The  slopes  of  the  bare 
lamp,  extensive  and  intensive  curves,  have  been  changed  by  multiplying  their  abscis- 
sae by  2). 

an  error  due  to  a  failure  of  the  inverse  square  law  itself.  This 
is  shown  by  all  the  curves  being  straight  lines  having  slightly 
different  intercepts.  There  may  be  an  appreciable  shifting  of 
the  effective  luminous  center,  but  there  is  no  appreciable  devia- 
tion from  the  inverse  square  law. 

Errors  in  illumination  calculations  due  to  an  error  in  the  dis- 
tance used  in  computations  is  given  by  curves  of  Fig.  3.  The 
error  due  to  assuming  the  distance — that  to  the  center  of  gravity 
of  the  light  source — is  probably  less  than  1  or  2  per  cent,  in  the 
usual  case  where  the  unit  has  been  photometered  at  the  distances 
of  about  10  or  12  feet   (3.048  or  3.657  m.). 

The  error  due  to  parallelism  must  be  small,  as  will  be  seen  by 


638 


TRANSACTIONS    I.    E.    S. PART    II 


reference  to  curves  of  Fig.  4.  These  show  the  effect  on  the 
graph  when  certain  percentages  of  the  illumination  at  10  feet 
(3.048  m.)  are  assumed  to  be  the  result  of  parallel  rays.  The 
deviation  from  a  straight  line  becomes  very  marked. 


8 
z 
27 

-5. 
o;Z6 

•-""4 
zo 

i-UUJ 

0P3 
K? 

0 

*-» 

1 

0 

<? 

x^ 

V*s» 

k* 

^K. 

2£s> 

S£/cg- 

s^ 

3     Z     4     6     8     10    12    14    16    18    20  22    24    26  28    30 
DISTANCE,  FEET 

Fige.  3. — Curve  showing  errors  in  illumination  computations  resulting 
from  error  in  the  distance. 


26 

*W 

on 

£22 

UJ 

|20 

§18 

<n 

§16 

§K 

UJ 

CC 
ot2 

5 

£8 
0 

£6 
4 
2 

0 

CUF 

IVE 

A- 

w\ 

fyp 

<f° 

tf$ 

tfl° 

)      2       4       6       8      10      12     14      16     18     20    22    24     26    28     50 
DISTANCE, FEET 

Fig.  4. — Effect  of  parallel  rays  on  measurements  of  illumination.  Curve  A  is  one  which 
would  be  obtained  from  the  point  source ;  other  curves  show  effect  of  parallel  rays 
for  given  percentages  of  illumination  at  10  ft.  distance,  due  to  parallel  component. 

The  results  of  this  investigation  seem  to  justify  the  conclu- 
sion that  for  all  practical  purposes  illumination  from  general 
lighting   units   may  be   computed   by   using   the    inverse   square 


EDWARDS   AND    HARRISON  :    ACCURACY   OF    PHOTOMETRY      639 

law  and  taking  the  distance  as  that  to  the  center  of  gravity  of 
the  light  source.  It  is  unnecessary  to  specify  the  distance  at 
which  distribution  curves  are  taken,  provided,  of  course,  that 
they  are  taken  at  a  distance  several  times  the  greatest  dimension 
of  the  unit.  Also,  it  is  unnecessary  to  use  the  term  apparent 
candle-power  in  connection  with  the  directional  intensity  of 
such  reflector  units  as  are  used  for  ordinary  illumination  pur- 
poses. 

III.  METHOD  OF  INVESTIGATING  ADJUSTMENT  ERROR,  AND 
THE  CALIBRATION  OF  PORTABLE  PHOTOMETERS. 

The  method  used  in  the  investigation  of  the  effect  of  parallel- 
ism in  the  photometry  of  reflector  units  suggested  itself  as 
being  a  very  good  one  to  apply  in  the  calibration  of  portable 
photometers.  In  fact,  in  the  previous  work  a  portable  photo- 
meter was  used  as  the  photometric  device,  and  the  bare  lamp 
test  originally  showed  the  photometer  to  be  out  of  adjustment. 
The  adjustment  error  was  corrected  by  means  of  the  data  ob- 
tained on  this  preliminary  run. 

Portable  photometers  have  a  variety  of  means  of  varying  the 
brightness  of  a  comparison  surface  which  is  matched  with  the 
surface  illuminated  from  the  test  end.  The  general  method,  here 
described,  is  applicable  to  all  when  consideration  is  given  to  the 
particular  principle  by  which  they  work. 

Consider  first  the  type  where  the  distance  from  the  compari- 
son lamp  to  the  diffusing  plate  is  varied,  such  as  the  regular 
Weber  photometer. 

Let  I1  =  c-p.  of  a  working  standard  which  can  be  placed  any 
distance  from  screen. 

I2  =  c-p.  of  comparison  lamp. 

dx  =  Distance  of  It  from  its  diffusing  plate. 

d„  =  Distance  of  I2  from  its  diffusing  plate. 

When  a  photometric  balance  is  obtained 

K  }    d?  '    d? 

where  /t  and  /,  are  respectively  the  transmission  coefficients  of 
the  diffusing  plates  for  lx  and  Ir 


640 


TRANSACTIONS    I.    E.    S. — PART    II 


—  is  a  constant  and  therefore  may  be  placed  equal  to  Kj  or, 

Since  the  scale  of  the  Weber  is  graduated  to  give  d2  directly, 
the  reading 

(3)   R  =  ^m'tt- 

Considering  Ix  and  I2  constant,  the  relation  of  R  and  d1  is,  of 
course,  the  equation  of  a  straight  line  passing  through  the 
origin. 


26 

111 

u-22 

cZ 
0 

g|2 

u-10 
0 

UJ  0 
O  0 

Z 

S6 
54 

2 

0 

■   -/ 

■"/ 

// 

A 

// 

mi 

>  y 

N0.1 

// 

A 

4t 

/f 

Vi 

'     2      4     6      8     10     12     rt     16     18    20    22   24    26    28    30 
PHOTOMETRIC  READINGS,  CENTIMETER5 

Fig.  5. — Adjustment  and  calibration  of  Weber  photometric. 

It  is  seen  from  equation  3  that  a  test  run  made  by  setting  up 
a  working  standard  lamp  at  various  distances  from  the  screen  of 
a  portable  photometer  serves  as  a  complete  test  of  the  accuracy 
of  adjustment  and  of  scale  graduation.  A  curved  or  irregular 
line  would  indicate  errors  such  as  incorrect  graduation  of  the 
scale.  The  failure  of  the  graph  to  pass  through  the  origin 
indicates  an  error  such  as  incorrect  position  of  the  comparison 
lamp.  Curve  1  of  Fig.  5  illustrates  the  manner  in  which  in- 
correct position  of  the  comparison  lamp  is  shown  by  this  method. 


EDWARDS   AND    HARRISON  I    ACCURACY   OF   PHOTOMETRY      64I 

Curve  2  of  Fig.  5  shows  the  results,  as  obtained,  after  correcting 
the  comparison  lamp  position  by  measuring  the  intercept  of 
Curve  1.  It  is  interesting  to  note  in  this  connection  that  the 
actual  position  required  to  give  the  zero  intercept  does  not  cor- 
respond exactly  with  the  measured  value,  due,  probably,  to  the 
reflection  of  light  on  the  inner  surface  of  the  tube. 

Letting  C  represent  the  multiplying  factor,  which  must  be 
applied  to  the  reciprocal  of  the  square  of  the  photometer  read- 
ings  to   obtain   illumination  values   in   foot-candles,   it   is   seen 

(R\ 2 
— ) , and 
1 

it   follows   that   the   constant   of   the   photometer   is   given   by 

multiplying  the  intensity  of  the  working  standard  by  the  square 

of  the  slope  of  the  graph. 

For  a  photometer,  such  as  the  Sharp-Millar,  where  the  read- 
ings are  made  proportional  to  the  reciprocal  of  the  square  of 
the  distance  to  the  comparison  lamp,  the  equations  are  changed. 

Here  the  relation  of  d.  and      . ■  should  be  a  straight  line  pass- 

y  R 

ing  through  the  origin.  Since  the  slope  of  the  line  in  this  case  is 
equal  to  ,  the  constant  of  the  photometer  is  given  by  mul- 

tiplying the  intensity  of  the  working  standard  by  the  square  of 
the  slope.  A  photometer  of  this  kind  is  usually  calibrated  to 
have  unity  constant,  so  this  method  does  not  lend  itself  particu- 
larly well  for  purposes  of  calibration,  but  is  very  useful  in 
testing  out  the  adjustment  of  the  instrument  as  well  as  the 
accuracy  of  a  particular  calibration.  In  the  laboratories  with 
which  the  authors  are  connected,  special  equipment  is  provided 
and  tests  of  this  kind  are  applied  to  all  portable  photometers  at 
frequent  intervals. 

The  change  of  constants  effected  by  the  use  of  absorbing 
screens  mav  also  be  accurately  determined  once  for  all  bv  this 
method.  If  there  are  a  number  of  absorbing  screens  which  have 
transmission  coefficients  mlt  tn2,  etc.,  which  can  be  used  either 
on  the  test  or  the  comparison  end  of  the  photometer,  the  new 


642 


TRANSACTIONS    I.    E.    S. PART    II 


constant  C  is  equal  to  —  if  a  screen  is  used  on  the    test  end 


7)1 


and  Cm  if  used  on  the  comparison  end.  The  change  in  con- 
stant is  then  obtained  by  a  calibration  test  with  and  without  an 
absorbing  screen,  and  the  transmission  coefficient  is  given  by 
the  ratio  of  the  two  constants  thus  obtained.  It  will,  of  course, 
remain  unchanged  except  as  affected  by  collection  of  dust  or 
dirt. 


IV.     ERRORS    IN    ILLUMINATION     MEASUREMENTS     DUE     TO 

FAILURE  OF  TEST  PLATE  TO  OBEY  THE  COSINE  LAW. 

In  measuring  the  illumination  in  the  usual  lighting  installation 

by  means  of  a  portable  photometer,  the  light  reaches  the  test 

plate  at  practically  all  angles.     A  test  plate  which  fails  to  obey 


S.9 


z.2 


i=0 


1       1 

[_ 

-CA 

LCI 

LAI 

ED 

FTOM 

COS 

1NE 

I  LAW 

1 

te; 

T  F 

ESI 

Lli 

1 

v- 

iv 

PEF 

<CE 

NT 

ERF 

!0R 

70 


60 


50, 


40  uj 


10 


0         10         20        30       40        50        60        70        80        90 

DEGREES  FROM  NADIR 

Fig.  6.— Error  in  photometric  test  plates,  due  to  failure  to  conform  to  the  consine  law. 


the  cosine  law  will  give  proper  weight  to  only  the  light  which 
strikes  it  at  an  angle  normal  to  it,  assuming,  of  course,  that  the 
photometer  has  been  calibrated  in  the  normal.  This  investiga- 
tion was  begun  after  noting  discrepancies  between  measured  and 
calculated  illumination  values  for  several  installations.  The  dis- 
crepancies were  noted  particularly  where  extensive-type  reflectors 
were  used. 

It  was  rather  surprising  to  learn  that  no  flat  plate  could  be 
found  which  did  not  show  a  very  considerable  error.  It  was  also 
rather  surprising  to  find  that  all  the  plates  tested,  although  ob- 


EDWARDS   AND    HARRISON  :    ACCURACY   OF    PHOTOMETRY      643 

tained  from  different  sources  checked  within  i  or  2  per  cent. 
The  complete  results  together  with  a  curve  showing  the  per  cent, 
error  are  given  in  Fig.  6.  The  illumination  values  are  given  for 
convenience  in  terms  of  ratio  to  the  normal. 


H 

Era 

E 

X 
2 

| 

"2.0         1.8         1.6         1.4         L2         LO         tt8        0.6        Q4 
SPACING  RATiaRATI0  =  -j^j|^7 

Fig.  7.— Average  per  cent,  error  in  illumination  values  for  an  installation  of  a  large 
number  of  units.   (Extensive  enamelled  steel  type  on  10  ft.  centers.) 


35°         25°       15°       5°  0  5°       15°       25°        35° 

Fig.  S. — Photometric  distribution  curve  of  bowl  shaped  enamelled  steel  reflector. 


[—16 

W 

gtz 

cc 
£8 

36 

a: 

1114 

2 

2 

0         L8         L6         1.4         1.2          1.0         08         0.6        0.4 
SPACING  RATIO.RATIO'^g1^ 

Fig.  9. — Average  per  cent,  error  in  illumination  values,  for  an   installation  of  a   large 
number  of  units.     (12. in.  mirrored  reflectors  on  10  ft.  centers. 

It  is  seen  that  the  error  at  450  is  more  than  10  per  cent.  The 
curves  of  Figs.  7,  9  and  11  are  the  result  of  laborous  computa- 
tions and  are  given  as  illustrations  of  the  magnitude  of  the  errors 
which  are  obtained  with  commonly  used  systems  in  large  rooms. 
Fig.  7  is  an  example  of  extensive  distribution  being  for  bowl- 


644 


TRANSACTIONS   I.   D.   S. — PART   II 


shaped  enamel  steel  having  the  distribution  shown  in  Fig.  8. 
Likewise,  Fig.  9  is  for  the  narrow  distribution  shown  in  Fig.  10. 
Fig.  11  is  still  another  example  showing  the  error  for  another 
type  of  extensive  distribution  shown  in  Fig.  12. 


25*       15"       5"  0    5"       IB- 
Fig.  io.—  Photometric  distribution  curve  and  mirrored  reflector. 


[— \b 

14 
gl2 

gio 

UJ 

<J6 
a; 

£4 

Z 

°2 

i 

0        L8        L6         M        LZ        LO        0.8        0.6        0. 

SPACING  RATiaRATO^PACING 

4 

Fig.  ii.—  Average  percent  error  in  illumination  values,  anfor  installation  of  a  large 
number  of  units.      (Fnamelled  steel  dome  shaped  reflectors  on  io  ft.  centers.) 


Fig.  12.— Photometric  distribution  curve  of  enamelled  steel  dome  shaped  reflector 

The  results  of  this  investigation  point  to  the  conclusion  that 
the  error  in  making  illumination  readings  with  the  average  flat 
test  plate  is  more  than  io  per  cent,  in  the  majority  of  cases.  The 
discrepancy  between  measured  and  calculated  values  which  have 


EDWARDS   AND    HARRISON  '.    ACCURACY   OF    PHOTOMETRY      645 

been  observed  are  fully  accounted  for.  For  accurate  work  a  care- 
fully designed  curved  surface  test  plate  should  be  used,  or  the 
proper  correction  should  be  applied  where  flat  plates  are  used. 

V.  METHOD  OF  OBTAINING  AND  RECORDING 
DISTRIBUTION  DATA. 

Distribution  curves  on  various  reflector  units  are  largely  for 


LAMP NO-  Q.W-L0 

VOLTAGE! K 3VLB£lear-J-30JL 

TOTAL  LUMENS  JQOO. 

LUMENS  PER  WATT-  !0- 
RUDUCTION  FACTOR  £W_. 


REFLECTOR 


DIAM..AL" DEPTH- 7. 

HOLDER  _  3=  tinker. 

TEST  ON 


TESTED  BYi CHECKED  BY.... 

APPROVED  BY.  -JiCALi 

PATE _ & il-il- ___ _  - 

REMARKS  SjroleCujyeJJniTS^Sini. 


ANGLE 

ZONE 

LUMEN! 

0 

::  •  5 

5 

m-.s 

0      10 

17.0 

15 

185-6 

0      20 

09.5 

25 

mo 

0     30 

1635 

35 

ms 

to 

?WS 

45 

less 

0     50 

ss 

1059 

0     56 

5Si-5 

65 

1404 

0      70 

»« 

75 

66-5 

0      60 

c'4o 

85 

70 

0      90 

Sii-S 

So 

is 

0    100 

105 

0    MO 

115 

0    IZO 

125 

0    1.30 

135 

0    iiCL 

1451 

0    150 

155 

0    160 

165 

0    170 

'175" 

0    180 

180 

c,r 

ZONE 

LUMENS 

v,^ 

W 

060 

J83.5 

S8-4 

c-,0 

■i" 

O-90 

Sir  5 

szt 

ISiO 

r:  >  ifir 

0.0 

O.t 

0-0 

* 

K1J-160 

00 

Oi 

6  0 

0-!60 

ez'.s 

su 

es3 

TO  MAINTAIN  UNIFORMITY  IN 
REC0RD1N6  DATA  THE  LAMP  USEPIN 
THISTEST  WAS  OPERATED  AT  AN  EFFI. 
aft  t  C1ENCY  OF  10  LUMENS  PER  WATTOR... 

A    W.RCTOREDUCETO W.RC 

'    MULTIPLY  CANDLE-POWER  AND 
LUMEN  VALUES  SC 


Clear  100  Watt  lamp 


Fig.  13- 


purposes  of  comparison  in  order  to  determine  the  most  suitable 
for  a  case  in  question.    The  fact  that  these  curves  have  been  made 


646  TRANSACTIONS    I.    E.    S. — PART    II 

with  lamps  operated  at  various  efficiencies  and  with  different 
reduction  factors  has  resulted  in  considerable  confusion  and  many 
erroneous  conclusions.  The  practise  has  been  to  obtain  and  put 
on  file  a  curve  for  each  kind  and  size  of  unit,  and  to  revise  the 
entire  sheet  each  time  thevlamp  efficiency  changes. 

A  new  method,  where  the  curves  are  obtained  by  using  a  lamp 
having  a  known  and  standard  reduction  factor  and  operated  at 
an  efficiency  of  10  lumens-per-watt,  has  been  devised.  A  sample 
curve  is  shown  by  Fig.  13.  It  is  intended  as  a  curve  which  is 
representative  of  a  type  of  reflector  and  which  can  be  applied  to 
various  sizes  and  with  lamps  of  various  efficiencies.  It  will  be 
noted  that  the  values  are  on  a  basis  of  1,000  total  lumens  for  the 
lamp,  which  is  very  nearly  correct  for  the  100-watt  lamp  at  1 
watt  per  horizontal  candle-power.  Assuming  the  standard  reduc- 
tion factor  78  per  cent,  for  the  100-watt  straight-sides  lamp,  the 
total  lumens  at  1  watt-per-candle  are  980,  or  2  per  cent,  less  than 
the  1,000  used.  On  the  basis  of  a  reduction  factor  of  81  per  cent., 
which  is  standard  for  the  larger  round  bulb  lamps,  the  total 
lumens  for  1  watt  per  horizontal  candle-power  would  be  1,018. 
Therefore,  the  specific  consumption  is  within  2  per  cent,  of  1 
watt-per-candle  for  both  the  straight-sides  and  round  bulb  lamps 
on  the  basis  of  10  lumens  per  watt,  and  the  average  happens  to 
be  within  0.1  per  cent. 

It  is  seen  that  all  values  of  lumens  are  instantly  convertible  to 
per  cent,  of  total.  The  intensity  values  are  as  easily  converted 
into  per  cent,  of  the  horizontal  for  the  bare  lamp,  with  an  error 
not  greater  than  2  per  cent.  The  original  sheet  is  provided  with 
a  blank  space  for  the  multiplying  factor  which  must  be  applied 
in  order  to  change  to  a  specified  efficiency,  the  intention  being  to 
fill  in,  on  blue  print  copies,  the  proper  factor  at  the  time  the 
sheet  is  sent  out.  Where  a  typical  curve  is  drawn  for  a  type  of 
reflector  for  all  of  its  sizes,  the  data  for  most  purposes  can  be 
corrected  by  multiplying  by  the  ratio  of  the  wattage  in  question 
to  100  watts.  A  further  correction  for  efficiency  can  be  made 
where  high  accuracy  is  desired. 

For  purposes  of  comparing  reflectors  no  reductions  are  neces- 
sary. The  comparisons  can  be  made  quickly  and  without  danger 
of  failing  to  take  into  account  the  differences  in  the  lamps  used. 


EDWARDS   AND    HARRISON  \    ACCURACY    OF    PHOTOMETRY      64/ 

It  has  been  found  that  the  number  of  curves  necessary  for 
ordinary  illumination  work  can  be  cut  down  to  a  small  part  of 
that  formerly  used,  and,  as  lamp  efficiencies  change,  it  is  only 
necessary  to  supply  a  new  multiplying  factor  to  bring  the  curve 
sheets  up  to  date. 

DISCUSSION. 

Mr.  T.  H.  Amrine:  With  reference  to  the  comparisons  of 
the  Bunsen  and  Lummer-Brodhun  photometer.  I  should  like  to 
ask  Mr.  Edwards  something  about  the  experience  of  the  operators 
in  this  test  with  the  Bunsen  photometer  as  compared  with 
the  Lummer-Brodhun,  whether  they  had  equal  experience  on  the 
two  photometers.  The  reason  I  ask  this  is  that  we  get  on  the  Bun- 
sen photometer  with  experienced  photometrists  somewhat  better 
accuracy  than  that  indicated  by  an  average  deviation  of  1.5  per 
cent.  For  instance,  I  have  the  result  of  tests  of  3  different 
operators  on  both  the  round  and  the  star  shaped  spots  from 
some  20  tests,  in  which  the  photometer  bar  has  masked  so  that 
the  operators  could  not  see  what  readings  they  were  getting. 
The  average  deviation  for  the  3  operators  was  0.497  per  cent. 
These  were,  of  course,  experienced  photometer  operators. 

Mr.  S.  L.  E.  Rose:  The  paper  just  given  by  Mr.  Edwards 
contains  a  great  deal  of  valuable  information  to  photometrists. 
The  part  of  the  paper  I  want  to  emphasize  is  that  of  making 
distribution  curves  using  a  100-watt  lamp  operated  so  as  to  give 
1,000  total  lumens.  This  one  curve  can  then  be  used  for  other 
reflectors  of  the  same  type  and  different  size  lamps  by  simply 
applying  a  factor.  The  factor  will  also  take  care  of  the  effici- 
ency at  which  the  lamps  are  to  be  operated  in  practise.  This 
will  probably  not  appeal  to  the  layman  as  usually  he  does  not 
want  to  bother  multiplying  by  any  factors ;  but  to  the  engineer 
it  should  appeal  very  strongly. 

Mr.  L.  J.  Lewinson  :  Some  of  us  are  probably  interested 
in  the  commercial  aspect  of  photometry.  It  might  be  in  order 
to  ask  Mr.  Edwards  about  the  relative  speed  attained  with  the 
Lummer-Brodhun  and  Bunsen  photometers.  Presumably  the 
time  reuired  to  make  a  measurement  is  longer  with  the  former. 
Does  the  increased  accuracy  of  the  Lummer-Brodhun  photo- 
7 


648  TRANSACTIONS    I.    E.    S. — PART    II 

meter  overbalance  the  extra  time  required  for  a  measurement, 
and  also  the  increased  eye  fatigue  which  probably  results? 

While  I  have  no  figures  which  are  directly  comparable  to  the 
results  shown  in  Fig.  I,  the  following  records  may  be  of  interest: 

At  the  Electrical  Testing  Laboratories  we  use  3  photometers 
equipped  with  Bunsen  screens  for  our  commercial  work.  As  a 
regular  routine,  about  20  per  cent,  of  the  lamps  measured  at 
each  photometer  are  checked  at  another  photometer.  One 
photometer  is  considered  as  the  standard,  and  the  variations  of 
the  measurements  made  at  the  other  photometers  computed. 
We  have  made  daily  averages  of  these  variations  over  long 
periods,  and  find  that  during  one  year  the  average  variation 
from  the  mean  value  was  about  1.17  per  cent,  and  for  the  follow- 
ing year  about  1.2  per  cent. 

Dr.  H.  E.  Ives  :  In  connection  with  this  discussion,  several 
years  ago  a  paper  came  out  by  Dr.  A.  E.  Kennelly  in  the  Trans- 
actions, of  the  National  Electric  Light  Association  entitled 
"Accuracy  in  Photometry."*  I  would  recommend  that  those  in 
terested  in  the  present  paper  to  look  it  up.  Several  of  the  ques- 
tions named  to-day  are  answered  in  it. 

Mr.  M.  LuckiEsh  :  This  paper  has  brought  out  some  par- 
ticularly interesting  points  and  the  measurements  have  settled 
some  doubts  in  my  mind  in  a  graphical  manner.  In  regard  to  dif- 
fusing screens,  the  best  that  I  could  make  or  otherwise  procure 
showed  errors  of  about  the  same  magnitude  as  those  examined  by 
the  authors.  Owing  to  the  failure  of  the  screens  to  be  perfectly 
mat  we  might  be  led  to  believe  that  this  is  always  a  serious 
matter.  In  a  room  with  dark  walls  the  illuminometer  measure- 
ments will  be  smaller  in  value  than  the  computed  measure- 
ments. However  in  all  practical  cases  there  will  be  some  light 
reflected  to  the  photometer  screen  by  an  indirect  route.  This 
will  tend  to  compensate  the  apparent  deficiency.  In  many  in- 
teriors this  additional  light  flux  which  is  not  computed  will 
amount  to  an  appreciable  fraction  of  the  total  light  flux  falling 
on  the  working  plane.  This  fault  of  diffusing  screens  is  there- 
fore not  a  very  serious  matter  in  most  cases.  The  authors 
have  illustrated  some   simple  procedures  which  might  well  be 

*  Proc,  National  Electric  X,ight  Association  (1908),  p.  208. 


ACCURACY  OF  PHOTOMETRY  649 

adopted  by  those  wishing  exact  knowledge  of  the  accuracy  of 
their  photometric  data. 

Mr.  E.  J.  Edwards  (In  reply)  :  Regarding  the  question 
brought  up  by  one  speaker,  I  would  say  that  the  inverse  square 
law  applies  even  to  very  concentrating  headlights  at  distances 
beyond  a  certain  point  up  to  which  there  is  likely  to  be  crossing 
of  the  rays  from  the  unit.  Attempts  to  carry  on  investigations 
along  the  lines  of  this  test  with  such  highly  concentrated  units 
as  ordinary  headlights  are  very  difficult,  due  to  lack  of  uniformity 
of  the  cross  section  of  the  beam.  It  is  very  hard  to  place  the 
photometer  to  catch  the  same  beam  at  the  various  distances. 
Tests,  so  far  as  they  have  been  carried  on,  seem  to  indicate  that 
the  inverse  square  law  holds  even  in  the  case  of  headlights  and 
other  projecting  units,  although  the  effective  distance  as  obtained 
by  a  graph  might  not  show  as  good  agreement  with  the  actual 
measured  distance  as  in  the  case  of  the  less  concentrating  type 
of  units  used  for  general  illumination  purposes. 

Mr.  Luckiesh  asked  as  to  the  effect  of  the  diffused  light  on 
the  test  plate  error  curves  in  any  assumed  system.  In  the  com- 
putations shown,  equally  spaced  stations  were  taken  and  the 
illumination  computed  from  a  distribution  curve  and  the  summa- 
tion was  carried  out  to  a  distance  away  from  the  unit  where  the 
increment  of  illumination  became  negligible.  This  method  takes 
no  account  of  the  diffused  light  from  walls  and  ceiling.  I  think 
it  would  be  very  hard  to  determine  accurately  just  how  much 
the  error  curve  would  be  affected  by  diffusion,  but  surely,  in 
the  cases  shown,  90  per  cent,  or  more  of  the  total  light  reaching 
the  test  plate  would,  under  average  conditions,  come  directly 
from  the  units,  and  therefore,  the  effect  on  the  curve  would  have 
to  be  very  small.  Even  the  diffused  light  component  would  have 
to  have  somewhat  the  same  error  since  it  reaches  the  test  plate 
from  all  angles. 

It  is  gratifying  to  the  authors  to  note  that  Mr.  Rose  empha- 
sized the  importance  of  considering  the  general  adoption  of  a 
simplified  means  of  handling  distribution  curves  such  as  the 
thousand-lumen  method  suggested  in  this  paper.  In  the  case  of 
a  curve  applying  specificially  to  a  250-watt  unit,  rather  than  to 


65O  TRANSACTIONS    I.    E.    S. — PART    II 

a  type  of  reflector,  it  would  probably  be  better  to  plot  on  the 
basis  of  10  lumens-per-watt  instead  of  1,000  total  lumens. 

In  reply  to  Mr.  Lewinson's  question  regarding  eye  fatigue  and 
the  time  taken  to  make  settings  with  the  Lummer-Brodhun  pho- 
tometer, I  would  relate  briefly  our  experience  in  changing  over 
our  equipment.  When  certain  operators,  who  had  been  accus- 
tomed to  using  the  Bunsen  screen,  were  asked  to  use  the 
Lummer-Brodhun,  they  complained  very  much  because  of  the 
fatigue  which  seemed  to  result  from  using  one  eye  instead  of 
two,  and  also  because  it  seemed  to  take  longer  for  settings.  It 
was  only  by  promising  these  operators  that  they  would  not  have 
to  use  the  Lummer-Brodhun  unless  they  preferred  them  at  the 
end  of  a  trial  period  that  we  were  able  to  get  them  to  willingly 
try  them  out.  At  the  end  of  a  week  they  said  they  were  able 
to  make  the  settings  more  easily  and  more  quickly,  although 
they  used  but  one  eye.  The  narrower  range  of  doubt  in  the 
settings  seemed  to  give  the  operators  a  greatly  increased  con- 
fidence in  their  work. 

In  this  connection  I  might  also  answer  Mr.  Amrine's  question 
regarding  the  experience  of  operators.  The  results  as  shown  in 
Fig.  1  were  for  an  operator  about  equally  experienced  with 
both  photometers.  There  were  other  results  obtained  in  this  test 
by  operators  who  had  been  accustomed  to  using  only  the  Bunsen 
screen,  but  since  a  fair  comparison  of  inherent  accuracy  of  the 
two  systems  was  desired  rather  than  the  accuracy  as  applied  to 
any  certain  class  of  operator,  they  are  not  included  in  Fig.  1. 
It  was  found  that  the  operators  who  were  unaccustomed  to  the 
Lummer-Brodhun  read  with  nearly  as  low  an  average  deviation 
as  the  experienced  operator,  but  were  incorrect  in  actual  value  be- 
cause they  set  according  to  some  appearance  of  the  field  other 
than  that  resulting  from  equal  intensity  on  both  sides.  I  agree 
with  Mr.  Amrine  that  it  is  possible  for  an  experienced  Bunsen 
operator  to  obtain  average  deviations  as  low  as  ^  per  cent,  as 
noted  by  him.  But  on  the  same  basis  we  believe  a  very  expert 
operator  under  most  favorable  conditions  might  show  less  than 
0.2  per  cent,  with  the  Lummer-Brodhun. 

The  flicker  photometer  was  not  considered  at  all,  since  it  had 
never  been  the  subject  of  thorough  investigation  by  us.     The 


ACCURACY  OF  PHOTOMETRY  65 1 

paper  is  not  intended  as  a  comprehensive  one  on  the  subject  of 
photometry. 

Another  speaker  has  shown  a  test  plate  error  curve  plotted 
on  the  same  basis  as  our  Fig.  6.  It  is  interesting  to  note  that 
his  tests  show  a  smaller  error.  Our  test  included  some  plates  of 
the  same  make  as  those  he  mentioned,  and,  although  they  showed 
a  somewhat  less  error  than  the  average,  they  came  out  greater 
than  the  result  he  showed.  All  of  the  plates  tested  showed  the 
same  characteristic  shape  of  error  curve.  We  did  not  find  that 
the  error  reached  a  maximum  at  a  certain  intermediate  angle.  It 
seemed  to  increase  in  a  smooth  curve  out  to  the  limiting  angle 
where  the  error  value  becomes  indeterminate.  Even  with  plates 
as  good  as  he  used  it  is  evident  that  test  plate  error  is  important, 
and  accounts  for  many  discrepancies  noted  in  the  past. 


652  TRANSACTIONS    I.    E.    S. — PART    II 

THE  STATUS  OF  THE  LIGHTING  ART* 


In  every  branch  of  human  endeavor  there  comes  a  time  when 
it  is  advisable  to  appraise  progress.  The  periodic  test  of  stu- 
dents' knowledge,  the  counting  of  cash  in  the  till,  the  taking  of 
stock  in  a  mercantile  establishment,  the  mariner's  observation, 
reconnaissance  in  military  manoeuvres, — are  all  recognitions  of 
the  need  for  measuring  progress  and  determining  status.  In  all 
forward  movements  such  determinations  are  admittedly  indis- 
pensable. 

The  science  and  art  of  illumination  have  made  considerable 
progress  in  the  recent  past.  The  growth  and  improvement  have 
been  rapid  and  of  a  nature  to  command  widespread  attention 
from  scientific,  commercial  and  humanitarian  viewpoints.  Ram- 
ifications are  so  numerous,  however,  and  the  variation  in  practise 
is  so  great  that,  even  in  the  opinions  of  those  best  qualified  to 
judge,  it  is  difficult  to  fix  the  present  status. 

To  possess  a  thorough  knowledge  of  the  status  of  the  lighting 
art  would  be  of  great  advantage.  First,  it  would  constitute  a 
record  for  future  reference  and  for  comparison  with  corres- 
ponding records  which  may  be  compiled  in  later  years,  and, 
second,  it  would  indicate  points  of  greatest  weakness  in  knowl- 
edge and  practise,  and  make  apparent  the  directions  in  which, 
because  of  such  weakness,  this  Society  has  the  greatest  oppor- 
tunity for  useful  service  in  the  near  future. 

Because  of  conviction  that  a  very  useful  purpose  would  be 
served  by  a  record  of  knowledge  of  the  principles  of  illumina- 
tion already  attained,  and  by  a  review  of  existing  practise  in 
the  art  of  illumination,  it  has  been  thought  advisable  to  under- 
take the  compilation  of  facts  in  an  endeavor  to  record  the  status 
of  the  lighting  art  in  the  United  States  of  America  in  the  year 
191 3.  The  obstacles  in  the  way  of  success  in  such  an  effort 
are,  however,  considerable.  The  range  of  variation  in  practise 
is  so  large  as  to  make  it  impracticable  to  present  a  really  com- 
prehensive survey.  More  -numerous  agencies  than  are  available 
are  required  in  order  to  obtain  sufficiently  reliable  and  consistent 

*  Presidential  address  by  Preston  S.  Millar  at  seventh  annual  convention  of  Illumin- 
ating Engineering  Society,  Pittsburgh,  Pa.,  September  22-26,  1913. 


THE    STATUS   OF    THE    LIGHTING   ART  653 

records  of  practise  in  various  parts  of  the  country  and  in  the 
several  branches  of  illumination.  To  present  a  fittingly  au- 
thoritative review,  a  more  accurate  and  detailed  knowledge  than 
the  writer  possesses  is  essential.  These  obvious  obstacles,  how- 
ever, were  deterrents  of  too  mild  a  nature  to  dissuade  from  an 
attempt  regarded  as  desirable. 

With  a  view  to  securing  information  regarding  lighting  prac- 
tise, questions  were  prepared  and  issued  about  the  first  of  July 
to  representative  individuals  and  commercial  organizations  of  the 
following  classes : 

Central  stations   (300  lists). 

Gas  Companies  (300  lists). 

Municipal  engineers  (200  lists). 

Ophthalmologists   (200  lists). 

School  associations  and  commissions  (200  lists). 

Street  railroad  companies  (200  lists). 

Manufacturers  of  incandescent  lamps  (in  various  quan- 
tities). 

Manufacturers  of  mantle  burner  lamps  (in  various  quan- 
tities). 

Manufacturers  of  arc  lamps  (in  various  quantities). 

Manufacturers  of  acetylene  supplies,  tips,  etc.  (in  various 
quantities). 

Manufacturers  of  oil  lamps  (in  various  quantities). 

Manufacturers  of  small  isolated  lighting  plant  equip- 
ments (in  various  quantities). 

Manufacturers  of  gasoline,  acetylene,  etc.  (in  various 
quantities). 

Manufacturers  of  lighting  glassware  (in  various  quanti- 
ties). 

Fixture  manufacturers    (in  various  quantities). 

Arc  lamp  post  manufacturers   (in  various  quantities). 

Railroads  (in  various  quantities). 

Street  lighting  companies  (in- various  quantities). 

The  country  was  divided  into  areas  of  approximately  1,000,000 
population  each,  and  so  far  as  practicable,  lists  of  questions 
were  sent  to  possible  sources  of  information  in  each  section  of 
1,000,000  population.    In  all  there  were  128  questions,  many  being 


654  TRANSACTIONS    I.    E.    S. — PART    II 

duplicates  of  questions  which  appeared  on  other  sheets.  Approxi- 
mately 1,750  lists  of  questions  were  issued.  The  questions  as  a 
whole  dealt  with  the  recognized  fundamentals  of  illuminating 
practise,  though  in  all  cases  the  attempt  was  made  to  adapt  them 
to  the  knowledge  which  the  correspondent  was  understood  to 
possess. 

It  was  anticipated  that  the  replies  to  these  questions  would  be 
relatively  few  in  number  and  that  the  information  thus  fur- 
nished would  be  inadequate,  both  in  respect  to  reliability  and 
comprehensiveness.  It  was  felt,  however,  that  in  the  aggregate 
the  information  elicited  would  be  interesting  and  valuable. 
Furthermore,  the  submission  of  these  questions  would  be  of  some 
value  in  attracting  attention  to  the  need  for  improvement  in 
lighting  practise,  and  in  drawing  attention  to  this  Society  as  the 
exponent  of  good  illumination. 

Approximately  20  per  cent,  of  those  receiving  the  questions 
promised  to  supply  such  information  as  they  possessed  or  could 
obtain,  and  approximately  one-half  of  these  have  been  heard 
from. 

In  discussing  the  subject,  it  is  necessary  to  adopt  some  form 
of  classification,  if  each  department  of  knowledge  and  practise 
is  to  be  considered  intelligently.  It  is  possible,  and  it  is  perhaps 
customary,  to  classify  with  respect  to  the  nature  of  the  installa- 
tion and  of  the  premises  illuminated.  It  has  seemed  preferable 
in  this  case  to  classify  according  to  those  features  which  deter- 
mine the  success  or  the  failure  of  the  illumination,  irrespective  of 
the  nature  of  the  installation.  Light  should  be  of  proper  quality 
and  the  lighting  equipment  should  be  installed  suitably.  In  the 
present  stage  of  our  knowledge  it  is  considered  that  illumination 
is  satisfactory  if  it  is  correct  in  regard  to  certain  qualities  of 
light  and  to  certain  features  of  utilization.  These  it  is  presumed 
to  present  to  you  as — 

The  Categories  of  Illumination. 

Light  Utilization 

Intensity  Contrast 

Direction  Cougruity 

Diffusion  Hygiene  and  Safety 

Color  Cost 
Steadiness 


THE   STATUS   OF   THE   LIGHTING   ART  655 

The  categories  are  believed  to  be  inclusive  of  all  ordinary 
features  of  illumination  which  should  receive  attention  from  the 
illuminating  engineer.  They  are  not  independent  of  one  another, 
but  are  interconnected  in  such  a  way  as  to  make  it  impracticable 
to  discuss  any  one  without  referring  to  one  or  more  of  the  others. 
It  is  also  impracticable  to  arrange  them  in  order  of  importance, 
since  their  relative  importance  varies  with  local  conditions  and 
with  the  requirements  of  the  installation. 

In  the  following  discussions  of  the  various  qualities  of  light, 
features  of  lighting  practise  and  correlated  matters,  the  informa- 
tion made  available  through  the  lighting  survey  which  has  been 
described  has  been  combined  with  such  material  as  could  be 
obtained  from  other  available  sources,  and  the  whole  is  com- 
pressed into  a  brief  review. 

INTENSITY   OF    LIGHT. 

Intensity  is  a  quality  of  illumination  which  has  received  very 
general  attention  from  the  earliest  days  of  illuminating  engineer- 
ing. Its  great  importance  was  recognized  early,  and  led  to 
study  and  to  the  development  of  photometers  for  facilitating  such 
study.  Early  writings  on  illumination  rarely  failed  to  include 
a  table  setting  forth  views  as  to  the  intensities  which  ought  to 
prevail  in  various  classes  of  installations.  It  is  true  that  these 
statements  were  very  largely  restricted  to  the  mean  horizontal 
illumination  at  some  height,  usually  30  inches,  above  the  floor, 
and  neglected  other  important  aspects  such  as  wall  brightness. 
Yet  emphasis  upon  intensity  undoubtedly  prompted  increase  of 
light  in  installations  where  it  had  been  inadequate,  and  tended  to 
raise  the  standard  everywhere. 

In  many  classes  of  lighting  the  intensity  standard  has  been 
increased  greatly  within  the  last  few  years.  Street  lighting,  very 
generally  inadequate,  has  felt  this  advance.  The  use  of  more 
powerful  illuminants,  the  growing  appreciation  of  importance  of 
lighting  business  streets  well,  and'  the  influence  of  merchants' 
display  lighting  systems  have  operated  to  increase  the  standard  of 
illumination  intensity.  In  the  middle  and  better  class  stores  a 
high  intensity  of  illumination  has  been  found  to  have  a  merchan- 
dising value,  and  the  intensity  standard  has  been  largely  increased 
in  the  past  few  years.     In  large  intelligently  conducted  industrial 


656  TRANSACTIONS    I.    E.    S. — PART    II 

plants  the  influence  of  improved  lighting  upon  output  and  upon 
safety  of  employees  is  evidenced  in  increased  light  intensity.  In 
small  factories,  as  in  small  stores,  this  improvement  is  less 
marked.  Sign  lighting  in  the  past  few  years  has  increased  greatly, 
indirectly  promoting  intensity  increase  particularly  in  street  light- 
ing. In  residences  the  advance  has  been  felt  perhaps  less  than 
elsewhere. 

The  attitude  of  manufacturers  of  illuminants  and  of  lighting 
companies  is  an  important  element  in  determining  the  trend  of 
practise  in  regard  to  intensities.  Manufacturers  of  gas  mantle 
lamps  have  been  active  in  promoting  the  use  of  their  product, 
displacing  open  flame  burners  with  large  increases  in  intensity. 
It  would  appear  that  there  exists  to-day  an  opportunity  for  gas 
companies  to  contribute  largely  to  the  improvement  of  illumina- 
tion by  more  actively  promoting  the  substitution  of  mantle  lamps 
wherever  open  flame  burners  are  still  in  use. 

In  the  electric  lamp  field  the  manufacturers  have  exploited  the 
Mazda  lamp  widely  and  their  efforts  have  been  seconded  more  or 
less  by  central  station  companies.  In  some  cities  the  central 
station  company  has  pushed  the  use  of  the  Mazda  lamp  actively 
with  very  beneficial  results.  In  other  cities  the  central  station 
company  has  not  taken  an  active  part  in  making  this  improvement 
in  electric  lighting  available  to  its  customers.  An  indication  of 
the  extent  to  which  the  Mazda  lamp  is  now  employed  will  be 
found  in  the  sales  records  of  the  electric  lamp  manufacturers, 
which  show  that  for  the  3  months  May,  June  and  July,  191 3, 
the  sales  of  the  several  types  of  lamps  have  been  as  follows : 

Per  cent. 

Gem  and  Carbon 37 

Mazda 63 

When  it  is  remembered  that  each  Mazda  lamp  lasts  two  or  three 
times  as  long  as  a  carbon  or  Gem  lamp,  it  will  be  seen  that  the 
use  of  Mazda  lamps  has  now  become  very  general,  especially 
when  it  is  recognized  that  in  general  it  is  the  more  extensively 
used  lamps  which  are  of  the  Mazda  type. 

The  sales  records  of  lamp  manufacturers  furnish  another  inter- 
esting evidence  bearing  upon  the  increase  in  the  light  intensity 
standard.  This  is  the  average  candle-power  of  the  lamps  sold 
during  recent  years.*    These  records  are  as  follows : 

*  Courtesy  of  General  Electric  Co. 


THE   STATUS  OF  THE  LIGHTING   ART  657 

Approximate  average  candle- 
power  of  all  incandescent 
Year  lamps  sold 

1906  18 

1907  19 

1908  21 

1909  23 

1910  25 

1911  26 

1912  29 

1913  (estimated)  32 


Standard  practise  in  regard  to  light  intensity  depends  upon 
conditions  other  than  simple  illumination  requirements.  The 
standards  vary  in  installations  of  similar  class  in  different  cities. 
For  example,  in  some  cities  merchants  have  not  come  to  appre- 
ciate the  advantage  of  adequate,  well  designed  illumination  to 
the  extent  that  they  have  in  other  cities.  Again  in  some  cities 
the  installation  of  "white  way  lighting"  has  been  found  to  mili- 
tate against  successful  show  window  lighting,  the  street  light  of 
higher  intensity  being  considered  sufficient  to  illuminate  the  show 
windows  for  ordinary  purposes.  In  other  cities  similar  installa- 
tions have  operated  to  increase  the  intensities  employed  in  show 
window  lighting,  it  being  found  that  in  contrast  to  the  lighting  of 
the  street,  more  light  is  necessary  in  order  to  make  the  window 
displays  as  prominent  as  they  were  when  the  street  lighting  was 
inadequate. 

In  few  classes  of  installations  is  it  practicable  to  measure  the 
advance  in  light  intensity  throughout  a  period  of  years.  While 
representative  data  are  available  for  modern  installations,  very 
generally  they  are  lacking  for  the  older  installations.  In  railway 
car  lighting  there  is  an  opportunity  to  measure  the  progress 
which  has  been  made,  due  to  the  fact  that  the  older  cars,  which 
have  antiquated  lighting  systems,  are  continued  in  service  on 
branch  lines  long  after  their  type  has  become  obsolete.  Minick* 
has  availed  himself  of  this  opportunity  to  place  upon  record  the 
average  illumination  intensities  which  are  typical  of  the  several 
systems  of  lighting  which  have  been  used  in  the  lighting  of  day 
coaches.    This  record  is  as  follows  : 

*  I.  E.  S.  Transactions,  May,  1913,  page  214  and  communication  to  the  writer. 


658  TRANSACTIONS   I.    E.    S. — PART   II 

Day  Coach  Lighting. 

Average  horizontal  foot- 
Installed  candles  36  inches 
during                            Description  of  illuininants  above  floor 

1850  to  1875     Oil  lamps — 2  wicks  feeding  one  flame 0.5 

— annular  wick 1.0 

1875  to  1900     Pintsch  gas  — 4  fish-tail  flames 1.5 

— 4  mantle  cluster 1 .65 

1880  to  1900     Carbureted  gasoline  — center  draft 1.3 

1900  to  date     Electric — 50-watt  opal  dip  lamps  and  flat  reflectors  1.4 

— 50-watt    clear   lamps   and   satin    finish 

bowl  reflectors 3.0 

Typical  illumination  intensities  for  artificial  lighting  of  certain 
classes  are  indicated  in  the  following  table: 

Typical  Intensities  of  Artificial  Illumination. 

Foot-candles 


Class  As  measured  through     Average       Usual  range 

Street  lighting — 

Principal  streets  in  cities-  -Horizontal  plane  of 

street  surface  ....       0.4  0.25-  2.0 

Important  side  streets Horizontal  plane  of 

street  surface 0.15         0.1  -  0.25 

Residence  streets Horizontal  plane  of 

street  surface  ....        0.04        0.01-  0.10 

Store  lighting Horizontal  plane  30 

inches  above  floor        4.0  2.0  -  6.0 

Show  window  lighting   Plan  of  trim 18.0  12.0-25.0 

Factory  lighting Horizontal  plane  30 

inches  above  floor        3.0  2.0  -  6.0 

Office  lighting Horizontal  plane  30 

inches  above  floor        3.0  2.0  -  4.0 

Residence  lighting  Horizontal  plane  30 

inches  above  floor        1.5  1.0-3.0 

Railway  car  lighting  ... Horizontal  plane  30 

inches  above  floor        2.0  1.0  -  3.0 

The  trend  in  illumination  intensities  is  upward.  Higher  ef- 
ficiency lamps  are  being  developed,  particularly  in  large  illu- 
minants  suitable  for  lighting  streets,  public  squares  and  large 
rooms  such  as  factories,  armories,  etc.  With  the  increase  in  ef- 
ficiency there  is  coming  into  our  practise  a  greater  insistence  upon 
good  candle-power  maintenance,  both  that  inherent  in  the  illumin- 
ants  and  that  secured  through  careful  maintenance  of  the  lighting 
equipment.  Daylight  illumination  in  interiors  is  perhaps  five  to 
ten  times  as  intense  as  that  provided  by  our  usual  artificial  light- 


THE   STATUS  OF  THE   LIGHTING   ART  659 

ing.  As  the  artificial  lighting  is  improved,  attaining  greater  merit 
in  respect  to  diffusion,  the  tendency  appears  to  be  to  increase  the 
intensities.  Latest  experiments  seem  to  indicate  that  it  is  not  im- 
probable that  when  we  shall  have  attained  more  complete  knowl- 
edge of  the  principles  of  good  illumination,  we  shall  find  that  the 
intensities  now  available  in  the  illumination  of  interiors  by  day- 
light will  have  to  be  approached  by  artificial  light  in  order  to 
satisfy  the  requirements.  It  is  apparent  further  that  as  appre- 
ciation of  the  beautifying  opportunities  in  illuminating  engineer- 
ing grows,  esthetic  considerations  impose  requirements  for  greater 
light  production.  Thus  the  requirements  of  ocular  hygiene  and 
of  esthetics  combine  in  demanding  the  production  of  more  light 
than  was  formerly  required  in  order  to  provide  acceptable  illu- 
mination. The  growing  appreciation  of  the  importance  of  good 
lighting  is  raising  the  standard,  making  a  still  further  general 
demand  for  the  production  of  more  light. 

DIRECTION  OF  LIGHT. 

Direction  of  light  has  not  received  the  same  extended  con- 
sideration as  have  some  other  features  of  the  illumination  prob- 
lem. It  is  forced  upon  attention,  however,  in  certain  classes  of 
work  where  improper  direction  brings  annoying  shadow  or  glare. 
In  such  cases  the  remedy  is  sometimes  found  by  changing  the 
direction  of  the  light. 

In  general,  artificial  lighting  is  provided  from  ceiling  fixtures  ill 
the  center  of  rooms  or  bays,  though  often  wall  brackets  are  used 
alone  or  to  supplement  center  lighting.  This  involves  a  down- 
ward direction  of  the  utilized  light.  Where  the  ceiling  is  em- 
ployed as  a  secondary  source,  as  in  indirect  lighting,  engineering 
thought  seems  to  favor  designs  which  will  very  largely  preserve 
a  downward  direction  for  the  light.  Daylight  illumination  of  in- 
teriors, on  the  other  hand,  is  very  generally  from  side  windows, 
and  the  lighting  has  a  strong  component  which  is  almost  hori- 
zontal, though  the  direction  is  usually  sjightly  downward. 

One  of  the  earliest  evidences  of  appreciation  of  the  importance 
of  proper  direction  in  lighting  is  the  well  established  tradition 
that  in  reading  "the  light  should  come  from  over  the  left 
shoulder."  Many  of  the  tenets  of  illuminating  engineering, 
though  unrecognized  in  the  formulation  of  this  precept,  are  evi- 


660  TRANSACTIONS    I.    E.    S. — PART    II 

dent  as  the  underlying  cause  which  has  led  to  its  wide  dissemin- 
ation and  general  acceptance.  It  was  the  outgrowth  of  reading 
experience  in  a  time  when  a  limited  amount  of  reading  was  done 
with  the  aid  of  a  single  illuminant  in  a  room.  This  homely  saying 
correctly  indicates  conditions  for  reading  which  avoid  shadow 
and  glare  and  best  contribute  to  the  ocular  welfare  and  comfort 
of  the  reader.  Under  the  conditions  of  use  which  obtained  when 
one  or  more  members  of  a  family  read  by  the  aid  of  light  from 
an  oil  lamp,  the  portability  of  the  light  source,  and  the  freedom 
of  the  reader  to  choose  any  desired  position,  made  compliance 
with  the  precept  entirely  practicable. 

Modern  conditions,  involving  immobile  light  sources  and  fixed 
positions  for  a  number  of  workers  in  the  same  room,  complicate 
the  problem  severely,  and  demand  much  more  adroitness  for  its 
successful  solution.  Indeed  were  proper  direction  of  light  the 
only  possible  solution,  as  is  implied  by  the  ancient  precept  just 
quoted,  the  problem  in  many  cases  would  be  very  difficult. 

Recently,  question  has  been  raised  as  to  the  general  propriety 
of  downward  light.*  Arguments  have  been  advanced  in  favor 
of  a  direction  of  light  which  is  from  the  side.  It  is  quite  possible 
that  in  the  near  future  developments  in  illuminating  practise  may 
result  in  less  general  adherence  to  a  downward  direction  for 
light  than  characterizes  present  practise. 

Meanwhile,  however,  the  general  downward  direction  of  light 
is  very  common.  In  store  and  office  lighting,  reflectors  designed 
to  redirect  the  light  downward  are  employed  very  generally. 
Only  15  to  25  per  cent,  of  stores  and  offices  employ  illuminants 
without  auxiliaries  of  some  kind.  In  many  of  these  exceptional 
installations  the  lamps  are  placed  so  near  the  ceilings  that  a 
general  downward  direction  of  light  is  obtained  in  spite  of  the 
absence  of  redirecting  auxiliaries.  Of  the  80  per  cent,  installa- 
tions which  employ  lighting  auxiliaries,  probably  four-fifths,  or 
about  65  per  cent,  of  all  installations  employ  some  device  designed 
to  direct  much  of  the  light  where  it  is  considered  to  be  most  use- 
ful. 

Under  daylight  conditions  the  direction  of  light  is  likely  to  be 
from  the  side,  and  altogether  too  little  attention  is  paid  to  adapt- 

*  Ives— Some  Home  Experiments  in  Illumination— I.  E.  S.  Transactions,  June,  1913, 
Page  229.  - 


THE    STATUS   OF   THE    LIGHTING   ART  66l 

ing  conditions  in  order  to  secure  the  best  illuminating  results. 
Indeed  where  a  number  of  persons  are  at  work  in  a  large  room 
illuminated  from  the  side  by  daylight,  it  is  well-nigh  impossible 
to  dispose  things  so  that  there  shall  be  practical  freedom  from 
shadows  and  glare. 

Direction  of  light  is  so  intimately  associated  with  diffusion  in 
contributing  to  the  merit  of  an  installation,  that  its  further  con- 
sideration may  well  be  included  under  the  following  caption. 

DIFFUSION  OF  LIGHT. 

Diffusion  of  light  in  earlier  discussions  of  illumination  prob- 
lems was  often  regarded  simply  as  a  quality  secured  as  a  result 
of  an  effort  to  conceal  light  sources  and  reduce  their  intrinsic 
brilliancy.  Its  influence  upon  the  appearance  of  a  room  and 
of  objects  in  a  room  was  recognized,  and  its  importance  from  the 
ocular  standpoint  was  often  considered,  though  not  fully  apprecia- 
ted. Importance  as  an  element  contributing  to  the  reduction  of 
sharp  shadows  completed  the  list  of  recognized  effects  of  diffusion. 
Larger  experience  in  illuminating  practise  and  more  recent  ex- 
periments and  research  have  brought  about  an  increased  and 
more  widespread  appreciation  of  the  importance  of  diffusion  as 
one  means  of  diminishing  glare  from  reflecting  surfaces  and 
promoting  ocular  welfare  in  general.  Increased  appreciation  of 
the  importance  of  securing  a  proper  degree  of  diffusion  is  perhaps 
the  most  notable  development  in  the  knowledge  of  illuminating 
engineering  during  the  past  two  or  three  years.  A  few  years  ago 
the  cry  was  for  the  use  of  efficient  reflectors  which  directed  the 
light  downward,  increasing  the  horizontal  illumination  intensity 
for  a  given  light  production,  or  decreasing  the  amount  of  light 
which  would  have  to  be  produced  in  order  to  provide  a  given 
illumination  intensity. 

At  this  time  such  reflecting  devices,  while  regarded  as  useful, 
are  recognized  as  inadequate  unless  some  means  of  providing  a 
fair  measure  of  diffusion  is  employed.  Thus  the  use  of  mat  re- 
flecting surfaces  is  growing.  One  year  ago,  it  was  found  that 
among  52  types  of  simple  glass  and  metal  reflectors  purchased 
upon  the  open  market  38  had  been  provided  with  etched  or 
other  diffusing  surfaces. 


662  TRANSACTIONS    I.    E.    S. — PART    II 

Where  diffusion  is  lacking,  multiplicity  of  sources  is  of  some 
assistance,  since  in  some  cases  glare  may  be  overcome  by  in- 
creased intensities  of  light  from  other  directions.  Multiplicity 
of  sources,  however,  means  multiplicity  of  shadows. 

Extremes  of  diffusion  are  rarely  required.  Too  much  diffu- 
sion means  characterless  illumination  and  may  mean  eye-fatigue. 
Too  little  diffusion  involves  glare.  The  degree  of  diffusion  which 
characterized  even  the  better  lighting  installations  of  a  few  years 
ago  is  now  recognized  as  inadequate.  "Semi-indirect"  and  "in- 
direct" lighting  succeed  in  many  installations  because  of  the 
higher  degree  of  diffusion  which  their  use  involves. 

Diffusion  of  light  cannot  be  separated  from  intensity  of  light 
in  discussions  of  this  character.  Obviously,  light  cannot  be  dif- 
fused without  loss.  Some  have  felt  that  with  diffused  light  a 
lesser  intensity  is  satisfactory.  Tests  have  been  reported  which 
purport  to  establish  this  point.  It  is  very  doubtful,  however, 
if  such  is  the  case.  With  a  correct  degree  of  diffusion  it  is  en- 
tirely practicable  to  determine  what  intensity  of  light  is  most  sat- 
isfactory for  a  given  purpose.  With  improperly  diffused  light 
it  has  been  found  that  the  same  intensity  of  light  is  unsatisfactory. 
This  has  been  taken  to  indicate  that  proper  diffusion  of  light 
results  in  a  decreased  intensity  requirement.  When  the  diffusion 
is  inadequate,  the  lighting  cannot  be  satisfactory  with  any  inten- 
sity provided  by  the  same  source  or  sources.  When  subjects 
who  are  being  experimented  upon  are  told  to  increase  the  intensity 
until  the  illumination  is  satisfactory  in  order  to  compensate  for 
the  lack  of  diffusion,  it  is  not  unnatural  that  the  result  should  be 
a  report  that  increased  intensity  compensates  for  the  lack  of 
diffusion,  giving  rise  to  the  conclusion  that  increased  diffusion 
permits  decreased  intensity.  It  is  submitted  that  such  an  effect 
has  not  been  demonstrated  and  there  is  no  reason  to  suppose 
that  insufficiency  of  diffusion  can  be  compensated  for  in  any  other 
way  than  by  increasing  the  diffusion.  Hence  the  intensity  re- 
quirement with  properly  diffused  light  cannot  be  said  to  be  less 
than  it  is  with  improperly  diffused  light.  Each  increase  in  light 
diffusion  has  brought  with  it  the  necessity  for  increasing  the  light 
production  in  order  to  compensate  for  the  loss  in  efficiency  in- 
volved in  the  diffusion. 


THE   STATUS  OF   THE   LIGHTING   ART  663 

In  one-half  to  two-thirds  of  the  total  installations  upon  which 
reports  have  been  obtained,  certain  degrees  of  diffusion  have 
been  obtained,  either  through  the  employment  of  diffusing  globes, 
by  depolishing  inner  surfaces  of  reflectors,  or  by  means  of  in- 
direct lighting. 

COLOR  OF  LIGHT. 

The  modification  of  light  to  produce  desired  colors  has  been 
very  generally  employed  in  recent  years  for  decorative  purposes. 
This  usage  has  been  very  largely  the  outgrowth  of  effort  intended 
to  secure  decorative  results.  The  ends  to  be  achieved  have  been 
more  or  less  clearly  defined,  but  the  means  to  be  employed  have 
in  general  received  too  little  attention.  Color  in  lighting  has 
been  the  agent  of  the  artist  rather  than  of  the  physicist  or  en- 
gineer. In  its  employment  empiric  rather  than  scientific  methods 
have  been  followed.  A  quality  of  light  which  must  ever  com- 
mend itself  to  the  artist  and  which  can  be  employed  effectively 
only  through  artistic  appreciation,  color  remains  both  a  physical 
and  a  physiological  phenomenon  which  must  be  applied  scien- 
tifically  to   secure   thoroughly   effective   results. 

The  art  of  illumination  seems  to  be  upon  the  verge  of  a  marked 
advance  due  to  appreciation  of  the  possibilities  of  color  manipu- 
lation in  lighting.  Illuminants  are  available  ranging  in  color 
from  the  neon  tube  through  the  Moore  tube  and  the  metallic 
electrode  arc  to  the  mecury-vapor  lamp  among  luminescent 
light  sources.  The  incandescent  light  sources  offer  a  narrower, 
though  material,  range  of  color  values.  The  employ- 
ment of  these  variously  colored  illuminants  in  conjunction  to  pro- 
duce color  effects  is  inherently  costly  and  troublesome,  and  likely 
to  be  practised  only  in  special  cases,  as  in  the  notable  installation 
in  the  Allegheny  County  Soldiers'  Memorial  Hall,  Pittsburgh.* 

The  modification  of  the  light  of  many  or  all  of  these  illumin- 
ants to  secure  light  of  desired  quality  is  a  field  which,  now  largely 
uncultivated,  offers  opportunities  of  which  the  illuminating  en- 
gineer is  beginning  to  avail  himself. 

The  use  of  differently  colored  illuminants  for  commercial  pur- 
poses is  beginning  to  be  appreciated.    In  a  few  dry  goods  stores, 

*  I.  E.  S.  Transactions,  Vol.  VI,  191 1,  Bassett  Jones,  Jr. 
8 


664  TRANSACTIONS   I.   E.    S. — PART   II 

small  area  lighting  by  illuminants  which  simulate  daylight  has 
been  provided.* 

For  many  purposes  we  aspire  toward  light  of  the  color  of 
daylight.  The  earliest  successful  illuminant  for  this  purpose, 
namely  the  Moore  carbon  dioxid  tube,  produced  light  of  practi- 
cally the  desired  quality.  Later  efforts  in  the  modification  of  light 
of  electric  incandescent  lamps,  incandescent  gas  mantle  lamps  and 
intensified  carbon  arc  lamps,  have  attained  some  measure  of  suc- 
cess. It  is  understood  that  all  these  forms  of  artificial  daylight 
have  come  into  use  to  a  limited  extent,  though  the  low  efficiency 
of  all  is  a  deterrent  preventing  their  application  upon  a  large 
scale. 

We  appear  to  be  upon  the  verge  of  developments  in  arc  lamps 
which  are  likely  to  result  in  relatively  high  efficiency  illuminants 
of  color  values  so  near  to  daylight  that  it  would  not  be  surpris- 
ing if  the  slighter  modification  of  light  which  would  be  entailed 
in  order  to  produce  artificial  daylight  would  be  accomplished  with 
a  loss  which  would  not  reduce  the  efficiency  of  the  lamps  below 
the  practicable  limit,  and  which,  therefore,  might  result  in  arti- 
ficial daylight  illumination  upon  larger  scales.  Such  a  consumma- 
tion is  to  be  desired,  because  artificial  daylight  will  not  only  be 
useful  in  a  practical  way,  but  experience  in  its  utilization  is 
likely  to  extend  the  realm  of  our  practical  knowledge  of  illumin- 
ating principles. 

Meanwhile  it  is  becoming  increasingly  apparent  that  for  certain 
purposes,  as  the  lighting  of  residences,  ballrooms,  etc.,  white  light 
is  not  acceptable  under  existing  conditions.  Having  become 
accustomed  for  generations  to  the  employment  of  light  in  which 
the  long  wave-lengths  are  more  accentuated,  we  have  either 
adapted  ourselves  to  that  condition,  or  there  is  some  quality  in- 
herent in  its  present  application  which  makes  it  more  satisfactory 
for  social  purposes  than  white  light.  The  amber  mantle  as  a 
substitute  for  the  ordinary  mantle  employed  in  gas  lighting  caters 
to  this  peculiarity  and  renders  the  gas  mantle  lamp  far  more 
acceptable  for  certain  purposes.  Among  electric  incandescent 
lamps  there  is  reason  to  believe  that  the  light  of  the  tungsten  fila- 
ment lamp  might  be  modified  to  advantage  in  order  to  produce  a 

*  For  example  see  Shalling— Store  Lighting,  I.  E.  S.  Transactions,  January,  1913. 


the:  status  of  the  lighting  art  665 

more  acceptable  color  for  social  lighting.  Attention  is  being  given 
to  these  matters,  and  it  is  reasonable  to  state  that  the  art  is  upon 
the  verge  of  developments  which  within  the  next  few  years  will  ex- 
tend employment  of  color  in  lighting,  and  that  lighting  auxiliaries 
will  be  developed  along  scientific  lines  which  will  yield  a  wide 
variety  of  tints  with  a  minimum  practicable  loss  in  efficiency. 

The  physiologist,  and  the  psychologist,  as  well  as  the  artist, 
are  interested  in  the  color  of  light.  Our  knowledge  of  these 
matters  is  not  great,  and  its  possession  is  not  general.  While  it 
appears  probable  that  we  are  upon  the  threshold  of  a  general 
effort  to  develop  the  use  of  color  in  illumination  along  scientific 
lines,  thereby  upholding  and  encouraging  artistic  endeavors  in 
this  direction,  it  remains  true  that  more  encouragement  should 
be  offered  to  the  study  of  the  effect  of  light  of  different  colors  upon 
the  eye  and  upon  the  mentality.  Only  as  artist,  psychologist, 
physiologist  and  engineer  combine,  will  the  employment  of  color 
in  illumination  be  developed  to  a  point  of  maximum  effectiveness. 
The  most  encouraging  fact  in  this  connection  is  the  activity  of  this 
Society  in  promoting  the  co-operative  study  of  this  problem  and 
in  disseminating  knowledge  accumulated  through  such  study. 

STEADINESS    OF    LIGHT. 

Steadiness  of  light  is  so  important  a  fundamental  that  most 
modern  illuminants  employed  in  indoor  lighting  are  free  from 
serious  objection  on  this  score.  Outdoors  the  require- 
ment for  steadiness  is  less  rigorous,  and  illuminants  pro- 
ducing light  which  is  relatively  unsteady  find  employment  more 
readily.  Among  such  are,  in  varying  degree,  arc  lamps  and  the 
flame  illuminants,  including  open  flame  gas  lamps,  and  some  types 
of  gas  mantle  lamps.  The  flame  illuminants  are  inclined  to  be 
unstable  when  exposed  to  wind.  In  modern  types  of  gas 
lamps  this  difficulty  is  reduced.  Among  arc  lamps,  the  flaming 
arc  is  particularly  unsteady.  Improvement  in  the  steadiness  in  the 
light  from  all  arc  lamps  is  effected  when  the  current  density  at 
the  electrodes  is  increased.  The  tendency  to  operate  these  lamps 
at  higher  currents,  as  noted  elsewhere,  brings  greater  stability 
in  addition  to  higher  efficiency. 

Cyclic  fluctuations  in  electric  lamps  constitute  another  form 
of    unsteadiness    of    light.      This    effect    is    the    more    notice- 


666  TRANSACTIONS    I.    E.    S. — PART    II 

able  in  the  case  of  arc  lamps.  Under  some  conditions  it 
has  been  found  practicable  to  operate  incandescent  lamps  upon 
25-cycle  alternating  current  circuits.  This  frequency  is  about  the 
lower  limit  and  under  some  conditions  the  flicker  of  lamps  so 
employed  is  objectionable.  Arc  lamp  operation  upon  25-cycle 
current  is  impracticable  where  good  lighting  is  the  desideratum. 
In  general,  steadiness  of  light  is  so  obvious  a  fundamental 
that  a  degree  of  unsteadiness  which  becomes  obtrusive  is  rarely 
tolerated. 

CONTRAST   IN   ILLUMINATION. 

Contrast,  here  regarded  in  a  comprehensive  sense,  is  a  most 
important  factor  in  the  utilization  of  light. 

Exposed  light  sources  within  the  field  of  vision,  occasioning 
too  great  contrast  with  the  surroundings  to  which  the  eye  is 
adapted,  are  detrimental  to  vision.  One  of  the  earliest  canons 
of  illuminating  engineering  branded  exposed  light  sources  as  the 
cause  of  ocular  discomfort  and,  under  extreme  conditions,  of 
impairment  of  vision.  The  development  of  more  powerful  and 
brilliant  light  sources  has  emphasized  the  importance  of  this  prin- 
ciple. Reduction  in  the  specific  intensity  of  light  sources  was 
recognized  as  a  means  of  mitigation.  Tables  of  specific  intensity 
of  various  light  sources  were  prominently  featured  in  the  earliest 
discussions  of  illuminating  engineering.  The  low  specific  inten- 
sity, for  example,  of  the  Moore  tube  was  urged  as  one  of  its 
chief  merits.  But  that  exposure  of  light  sources  was  not  gen- 
erally recognized  as  a  serious  menace  to  eyesight  and  a  source  of 
discomfort,  is  attested  by  the  designs  of  lighting  fixtures  and 
glassware  of  the  period.  Catalogues  of  lighting  equipment  fur- 
nished by  the  leading  manufacturers  at  that  time  show  designs 
which  in  the  light  of  our  present  practise  appear  almost  barbarous 
in  their  defiance  of  this  first  principle  of  illuminating  engineering. 

The  evil  effects  of  exposed  light  sources  include,  as  previously 
mentioned,  ocular  discomfort  and,  under  certain  conditions,  im- 
pairment of  vision  in  the  sense  both  of  injury  to  eyesight  and 
diminished  visual  power.  These  effects  in  rather  loose  termin- 
ology have  been  attributed  to  glare.  A  number  of  causes  and 
effects  differing  materially  in  character  have  been  classed  as 
glare  without  proper  distinction,  due  to  their  nature.    The  study 


the;  status  of  the;  lighting  art  667 

of  glare  and  its  evil  effects,  insofar  as  these  are  concerned  with 
the  exposure  of  light  sources,  received  a  notable  impetus  in  the 
early  years  of  this  Society's  history.  This  was  particularly  true 
of  glare  as  an  agent  which  diminishes  visual  power.*  Following 
upon  the  discussion  of  glare  in  the  sense  here  considered,  there 
came  a  more  general  appreciation  of  the  advantages  of  concealing 
light  sources.  Leading  manufacturers  of  lighting  appliances  dis- 
continued some  of  the  most  prominent  offending  designs  and  the 
newer  designs  evidenced  more  attention  to  this  important  aspect 
of  the  problem.  While  on  all  sides  one  finds  evidence  of  continued 
neglect  of  this  important  fundamental,  yet  we  may  feel  that,  at 
least  by  lighting  practitioners,  there  is  a  thorough  understanding 
of  the  serious  nature  of  exposure  of  light  sources,  and  that  a  real 
effort  is  being  made  to  eliminate  it  from  our  practise.  It  is  now 
generally  recognized  that  entire  concealment  of  the  light  source 
as  in  indirect  lighting,  reduction  of  its  brilliancy  by  the  aid  of 
diffusing  media,  as  in  much  of  the  direct  and  "semi-indirect" 
lighting,  or  the  removal  of  the  light  source  from  the  ordinary 
field  of  view,  is  an  essential  to  good  lighting.  Very  generally 
these  principles  are  being  put  into  practise.  Flat  reflectors,  which 
exposed  the  light  source,  are  being  replaced  by  bowl  reflectors 
which  conceal  it  from  ordinary  view ;  manufacturers  of  gas 
lamps  report  growing  use  of  diffusing  rather  than  clear  globes; 
manufacturers  of  electric  lamps  report  increasing  use  of  frosted 
lamps ;  the  use  of  indirect  and  "semi-indirect"  lighting  fixtures  is 
growing  rapidly ;  new  installations  of  street  lamps  for  civic  light- 
ing are  very  generally  mounted  higher  than  was  the  practise  a 
few  years  ago;  the  proportion  of  bare  lamp  installations  is  de- 
creasing. 

Exposed  light  sources  of  the  more  usual  types  are  more  bril- 
liant than  is  the  sky,  and  surroundings  are  apt  to  be  less  bright 
under  artificial  light  than  they  are  in  the  daytime.  Thus  greater 
contrasts  prevail  in  artificial  lighting.  In  the  better  lighted  instal- 
lations, however,  improvement  is  now  being  effected  both  through 
the  reduction  of  the  specific  intensity  of  the  sources  and  through 
the  increase  in  the  brightness  of  the  surroundings. 

Glare  from  reflecting  surfaces  due  to  specular  reflection  is 
occupying  much  thought  of  investigators  and  practitioners  at  the 

*  A.  J.  Sweet,  Journal  Franklin  Institute,  May,  1910. 


668  TRANSACTIONS   I.    E.    S. — PART    II 

present  time.  This  again  involves  the  question  of  contrast. 
Indeed,  all  manifestations  of  glare  are  but  little  more  than  im- 
proper contrast.  A  printed  page  of  glossy  paper  may  give  rise 
to  a  serious  condition  of  glare  due  to  specular  reflection.  This 
means  that  one  views  the  imperfectly  reflected  image  of  a  light 
source  which,  in  contrast  with  the  surroundings,  is  of  excessive 
brightness.  The  result  is  ocular  discomfort.  If  glossy  ink  has 
been  used  in  printing,  it  may  be  difficult  or  even  im- 
possible to  discern  the  imprint.  This  is  a  case  in  which  the 
excessive  specular  reflection  which  causes  discomfort  by  too  great 
contrast  with  the  surroundings  is  complicated  by  too  little  local 
contrast  to  permit  of  reading  the  glossy  black  letters,  which  reflect 
specularly  almost  as  well  as  the  surrounding  surface  of  the  paper. 
This  element  of  contrast  is  receiving  a  fair  share  of  the  atten- 
tion which  it  merits,  and  our  knowledge  of  the  principles  involved 
is  growing  rapidly.  This  is  followed  by  improvement  in  the 
design  of  new  installations  in  which  such  knowledge  is  applied. 
In  general,  therefore,  it  may  be  said  that  improper  contrast  which 
manifests  itself  as  glare  either  due  to  exposed  light  sources  or  to 
reflections  from  specular  surfaces  is  rapidly  being  brought  under 
control  in  the  installations  which  are  receiving  illuminating  engi- 
neering attention.  The  lesser  degrees  of  difficulties  involved  in 
improper  contrast  are,  however,  not  so  thoroughly  understood  or 
appreciated;  our  knowledge  appears  to  be  in  need  of  con- 
siderable extension  and  our  practise  appears  to  suffer  from  lack 
of  proper  care  for  this  important  feature  of  the  utilization  of 
light. 

CONGEUITY. 

The  past  2  or  3  years  have  witnessed  a  marked  increase  in  gen- 
eral appreciation  of  the  importance  of  the  esthetics  of  illumina- 
tion. It  is  becoming  generally  recognized  by  lighting  practitioners 
that  lighting  equipment  as  well  as  the  illumination  produced  must 
be  in  harmony  with  the  character  of  the  premises  which  are 
illuminated. 

In  the  early  development  of  the  science  and  art  of  illumination 
it  was  but  natural  that  the  physical  and  engineering  aspects  should 
receive  first  attention,  for  these  underlie  the  entire  art.     It  was 


THE  STATUS  0E  THE  LIGHTING  ART  I  69 

but  natural  that  first  developments  should  evidence  incongruities. 
This  period  of  illuminating  engineering  is  now  being  outgrown. 
The  art  is  now  approaching  the  period  of  adolescence  in  which 
the  acquirement  of  knowledge  is  rapid  and  preparation  for  mature 
effort  is  the  keynote.  At  this  time  it  is  being  appreciated  that 
congruity  in  the  utilization  of  light  in  any  installation  is  a  desid- 
eratum of  first  importance.  It  is  just  as  inartistic  to  locate  highly 
embellished  ornate  and  inefficient  fixtures  and  glassware  in  a 
machine  shop,  as  to  hang  tin  shades  in  an  elegantly  furnished 
drawing  room.  Error,  however,  occurs  commonly  in  the  use  of 
inartistic  lighting  equipment  where  artistic  design  is  required, 
and  it  is  in  this  aspect  of  general  practise  that  one  of  the  great 
needs  for  improvement  is  apparent. 

It  is  recognized  that  the  illuminating  engineer  should  be  con- 
versant with  and  appreciative  of  art  in  its  several  phases  if  his 
work  is  to  reveal  in  an  intelligent  sympathetic  manner  the  best 
of  the  design.  Some  instances  are  encountered  where  only  the 
artist  can  do  justice  to  the  requirements.  In  such  cases  the 
illuminating  engineer  should  be  guided  by  the  artist,  and  he  should 
be  able  to  accomplish  the  desired  end  more  effectively  and  with  a 
lower  cost  than  the  artist  can.  The  latter  fully  appreciates  what 
he  desires,  but  presumably  has  not  familiarized  himself  with 
lighting  technique  to  the  extent  which  enables  him  to  accomplish 
his  purpose  so  well  as  can  the  illuminating  engineer.  It  is  believed 
that  the  events  of  the  past  2  or  3  years  have  contributed  toward  a 
somewhat  better  mutual  understanding  and  that  in  the  very  near 
future  we  shall  enjoy  a  fuller  measure  of  the  needed  hearty 
co-operation  of  architects,  decorators  and  fixture  designers, 
thereby  achieving  improvements  in  the  artistic  phases  of  illumina- 
tion which  are  so  much  needed  at  the  present  time. 

In  this  connection  it  may  be  noted  that  a  most  important  factor 
in  the  betterment  of  illumination  conditions,  artistically,  is  im- 
provement in  the  design  of  stock  fixtures  and  stock  lighting 
auxiliaries.  Any  given  type  of  fixture  must  of  course  be  designed 
for  one  set  of  conditions  and  cannot  be  expected  to  be  congruous 
in  a  wide  variety  of  installations.  It  may,  however,  be  made 
tasteful  in  itself,  free  from  objectionable  features  and  well 
adapted  for  general  use  under  average  conditions  of  the  class  for 
which  it  is  designed.     When  considering  the  artistic  aspects  of 


670  TRANSACTIONS    I.    E.    S. PART    II 

illumination,  it  is  not  well  to  restrict  thought  to  individual  designs 
which  of  course  are  expected  to  surpass  stock  fixtures  being  at 
once  more  pleasing  and  congruous  when  applied  in  the  installa- 
tions for  which  they  are  designed.  It  should  be  remembered  that 
a  5  per  cent,  improvement  in  stock  fixtures  affecting  large  num- 
bers of  people  may  be  of  much  greater  value  than  a  100  per  cent, 
improvement  in  a  distinctively  designed  installation. 

Residences  and  certain  other  classes  of  installations  may  be 
said  to  be  equipped  very  generally  with  fixtures  which  cannot  be 
approved  either  from  the  artistic,  hygienic  or  efficiency  stand- 
points. This  is  particularly  true  of  installations  which  date  back 
5  years  or  more.  It  is  especially  true  of  combination  fixtures 
designed  to  permit  the  use  of  either  electric  or  gas  illuminants. 
The  presence  of  these  fixtures  which  are  neither  tasteful  in  them- 
selves nor  consistent  with  the  standard  of  good  taste  exhibited  in 
the  furnishing  of  rooms  in  the  middle  and  better  class  residences, 
constitutes  a  barrier  to  the  improvement  of  lighting  conditions  in 
residences.  The  first  step  toward  a  general  improvement  in  such 
lighting  conditions  would  appear  to  be  the  wholesale  replace- 
ment of  such  fixtures.  Such  is  the  crying  need  of  the  time.  The 
way  of  accomplishing  this  is  not  clearly  indicated.  The  extension 
of  lighting  improvements  to  residence  and  to  some  other  classes 
of  lighting  is  dependent  upon  the  displacement  of  such  fixtures. 

It  is  fortunate  that  displacement  of  incongruous  lighting  equip- 
ment by  equipment  which  would  be  suitable  from  the  artistic  or 
efficiency  standpoint  would  bring  about  automatically  an  im- 
provement in  installations  from  the  standpoint  of  ocular  welfare. 
If  the  requirements  for  artistic  and  hygienic  equipment  were 
diverse  or  incompatible,  the  situation  would  indeed  be  difficult. 
As  it  is,  the  happy  concordance  of  hygienic  and  artistic  require- 
ments constitutes  a  demand  for  betterment  which  cannot  be 
ignored. 

HYGIENE  AND  SAFETY. 

In  ordinary  lighting  practise  the  qualities  of  illuminants  are 
such  that  laws  of  hygiene  are  rarely  transgressed.  Much  has 
been  written  regarding  unsanitary  effects  of  various  illuminants. 
but  usually  it  has  been  found  that  such  articles  are  the  work  of 
the  press  agent  of  some  manufacturer  or  promoter  of  rival-  il- 


THE   STATUS   OF   THE)    LIGHTING    ART  67 1 

luminants.  Ordinarily,  ventilation  is  sufficiently  good  to  avoid 
deleterious  effects  from  noxious  gases.  Most  artificial  illuminants 
are  not  sufficiently  strong  in  ultra-violet  light  to  be  injurious. 
The  quartz  mercury-vapor  lamp  perhaps  forms  an  exception, 
but  the  heavy  glass  globe  which  is  invariably  used  with  it  when 
employed  for  lighting  purposes  is  sufficient  protection. 

On  the  other  hand,  the  use  of  light  promotes  sanitation  in 
securing  greater  cleanliness  wherever  it  is  applied.  Likewise  the 
more  liberal  and  judicious  use  of  light  promotes  safety.  This  is 
being  recognized  in  the  industries  among  the  larger  and  more 
progressive  corporations,  and  more  special  attention  is  being 
devoted  to  promoting  safety  by  the  use  of  light. 

Of  much  greater  importance  than  general  hygiene  is  ocular 
hygiene  in  its  relation  to  illumination.  It  is  not  from  some  delet- 
erious quality  of  the  light  of  a  particular  kind  of  illuminant  that 
harm  to  the  eyes  results,  but  rather  from  the  misuse  of  light,  irre- 
spective of  the  illuminant.  The  efforts  which  are  now  being 
made  to  promote  the  correct  use  of  light,  particularly  in  the 
home  and  in  schools,  are  of  incalculable  advantage  to  the  public 
in  safeguarding  especially  the  eyes  of  children  during  the  period 
of  immaturity  when  they  are  more  susceptible  to  the  ill-effects 
of  misuse  of  light. 

Ocular  hygiene  is  being  investigated  in  a  number  of  labora- 
tories. Generally,  conditions  of  visibility  are  judged  by  means  of 
a  determination  of  the  threshold  visibility  value.  This  involves 
the  determination  of  either  minimum  light  intensity  for  visibility, 
minimum  size  of  object  viewed,  minimum  contrast  which  can  be 
perceived,  or  of  the  time  element  in  the  perception  of  objects  at 
the  threshold  value.  While  it  perhaps  remains  to  be  determined 
how  far  the  results  of  such  investigations  may  be  considered 
applicable  to  practical  lighting  conditions,  yet  it  is  undoubtedly 
a  fact  that  the  information  which  is  being  made  available  as  the 
result  of  such  investigations  is  advancing  the  science  of  ocular 
hygiene  and  contributing  largely  to  knowledge  of  the  principles 
of  good  lighting.  For  the  further  promotion  of  ocular  welfare 
there  is  need  of  further  research  in  which  the  combined  efforts 
of  illuminating  engineers,  ophthalmologists  and  psychologists 
should  govern  the  nature  of  the  investigation.    Also  there  is  need 


6/2  TRANSACTIONS    I.    E.    S. PART    II 

of  the  application  of  the  results  of  such  investigation  to  practical 
lighting  conditions.  Conditions  appear  to  be  such  as  to  warrant 
the  assumption  that  the  next  few  years  will  witness  considerable 
advances  along  both  these  lines. 

COSTS. 

It  is  usually  considered  that  the  cost  of  artificial  lighting 
includes — 

Cost  of  lighting  equipment. 

Cost  of  maintaining  equipment  including  interest,  depre- 
ciation, etc. 

Cost  of  fuel  or  energy  required  for  operating  the  system. 
Cost  accounting  in  illumination  work  is  well  handled  in  larger 
organizations  and  possible  efficiency  improvements  are  therefore 
considered  intelligently.  Much  lighting,  however,  is  not  organ- 
ized and  the  expense  is  borne  by  those  who  do  not  handle  accounts 
intelligently  and  who  are  not  in  a  position  to  judge  of  the  ulti- 
mate efficiency  which  may  be  obtained  in  lighting  by  the  employ- 
ment of  the  several  available  systems.  Thus  the  use  of  open 
flame  gas  burners  and  of  carbon  and  Gem  lamps  continues  in 
many  installations  where  the  more  efficient  mantle  burners  and 
Mazda  lamps  should  be  used. 

Artificial  lighting  is  very  inefficient,  due  first,  to  the  low  effi- 
ciency of  energy  transforming  devices,  and  second,  to  the  low 
light  production  efficiency  of  illuminants.  To  this  in  many  cases 
must  be  added  unintelligent  utilization  of  light.  In  spite  of  its 
low  efficiency,  however,  the  cost  of  artificial  lighting  is  small. 
The  following  lighting  costs  are  suggestive  in  this  connection : 

Approximate  cost  of  artificial  lighting 
Class  of  lighting  installation  in  proportion  to  total  operating  costs 

Small  wage  earner's  home — ratio  of  cost 

of  lighting  to  total  income I  per  cent. 

Well  conducted  large  manufacturing  es- 
tablishments   which    are    well    lighted 

with  modern  illuminants — ratio  of  cost  of 

artificial  lighting  to  total  cost  of  output 

exclusive  of  selling  expenses J  3  to  *4  of  1  per  cent. 

Large   retail   mercantile  establishments — 

ratio  of  total  lighting  cost  to  total  sales    Probably  less  than  i.oper  cent. 
Small  stores — ratio  of  total  lighting  cost  to 

total  sales 2  per  cent. 

Modern  loft  buildings 1  to  2  per  cent. 


THE  STATUS  OF  THE  LIGHTING  ART  673 

Street  lighting  appropriations  by  municipalities  are  of  the  order 
of  60  cents  to  $1.00  per  inhabitant  per  year. 

In  comparison  with  the  benefits  conferred  by  artificial  lighting 
both  in  the  way  of  added  commercial  advantage  and  extended 
opportunities  in  education,  social  life  and  recreation,  the  cost  of 
artificial  lighting  is  remarkably  low.  In  many  cases  where  its 
beautifying  influence  is  of  paramount  importance,  the  cost  is 
immaterial. 

In  this  country  there  is  a  strong  sentiment  in  favor  of  corporate 
work  in  the  promotion  of  employees'  welfare.  The  tendency 
of  the  American  business  man  toward  organization  and  scientific 
management  was  never  stronger  than  at  the  present  time.  From 
both  viewpoints,  artificial  lighting  is  an  important  factor  and  in 
both  respects  its  cost  is  small  as  compared  with  the  advantages 
derived. 

The  cost  of  artificial  lighting  has  been  rapidly  reduced  as  a 
result  of  the  development  of  improved  illuminants  and  of  the 
greater  economies  which  have  been  effected  in  manufacture  and 
operation.  Doubtless  it  will  be  still  further  reduced  in  the  near 
future.  This  reduction  in  cost  has  been  largely  automatic  and  the 
outgrowth  of  economic  conditions. 

With  the  reduction  in  the  cost  of  lighting  has  come  betterment 
in  lighting  conditions.  Much  that  is  advanced  in  the  lighting 
practise  in  this  country  is  due  to  the  progressive  enterprising 
manner  in  which  manufacturers  of  illuminants  and  lighting  com- 
panies have  conducted  their  businesses.  In  the  future  we  must 
look  to  such  organizations  to  bring  about  improvements  in  light- 
ing practise  which  are  needed  if  the  work  of  this  Society  is  to 
be  applied  for  public  benefit. 

Only  to  an  extent  so  small  that  it  is  negligible  has  the  decrease 
in  the  cost  of  lighting  been  effected  through  legislation.  Recently, 
however,  in  a  few  cities  recourse  has  been  had  to  rate  legislation 
affecting  public  service  corporations  .in  consequence  of  which  rela- 
tively low  rates  have  been  imposed.  These  low  rates  imposed 
by  city  regulation,  actuated  often  by  political  rather  than  by 
economic  motives,  constitute  a  menace  to  the  success  of  this 
Society's  work  insofar  as  it  may  affect  the  general  public.  Only 
as   corporations   engaged   in   supplying  lighting   service   earn  a 


674  TRANSACTIONS   I.    E.    S. — PART    II 

fair  return  upon  their  investments  can  they  be  expected  to  con- 
tinue operation  along  broadly  progressive  lines.  Only  as  such 
policies  shall  prevail  among  manufacturers  of  lighting  appliances 
and  in  the  management  of  lighting  companies,  may  we  hope  for 
rapid  progress  in  the  further  improvement  of  lighting  conditions. 
Commercial  organizations  have  great  potentialities  for  good,  and 
we  can  rely  upon  a  continuance  of  their  assistance  only  if  the 
further  reductions  in  the  cost  of  lighting  are  effected  in  accord- 
ance with  the  development  of  the  economic  situation  rather  than 
in  accordance  with  the  artificial  conditions  of  political  regulation. 

The  foregoing  brief  review  is  based  in  part  upon  information 
which  has  been  made  available  by  replies  to  the  questions  which 
were  issued  as  a  part  of  the  lighting  survey.  These  replies  are 
extensive  and  numerous,  affording  in  the  aggregate  a  large  amount 
of  statistical  information.  They  apply,  however,  to  so  small  a 
percentage  of  the  total  lighting  industry  as  to  make  it  seem  unwise 
to  present  a  complete  summary  of  the  facts  which  they  make 
available.  All  the  information  which  has  been  obtained  in  this 
way  is  available  to  the  society  and  may  be  very  useful  in  the 
furtherance  of  statistical  work  along  this  line. 

In  the  preparation  of  the  detailed  information  which  has  been 
made  available  as  the  result  of  this  lighting  survey,  the  writer 
has  been  favored  by  the  kind  co-operation  of  individuals  and 
companies  so  numerous  as  to  make  it  impracticable  to  list  them 
in  this  connection ;  it  is  possible  only  to  express  deep  apprecia- 
tion. The  loyal  disinterested  assistance  of  Mr.  N.  D.  Macdonald 
of  the  Electrical  Testing  Laboratories  should  be  recorded 
especially. 

CORRELATED   MATTERS. 

In  addition  to  a  review  of  knowledge  and  practise  in  respect 
to  the  several  qualities  of  light  and  features  of  utilization,  no  dis- 
cussion of  the  status  of  the  lighting  art  would  be  complete  with- 
out reference  to  a  number  of  supplementary  factors  which  have 
been  influential  in  establishing  the  present  status  and  which  may 
be  looked  to  for  assistance  in  advancing  that  status.  Prominent 
among  these  are  educational  agencies,  commercial  organizations 
in  the  lighting  field,  the  Illuminating  Engineering  Society,  the 
attitude  of  related  professions  and  industries,  and  photometry. 


THE   STATUS  OF  THE  LIGHTING  ART  675 

EDUCATIONAL  AGENCIES. 

In  the  way  of  technical  education,  universities  and  colleges 
are  devoting  more  attention  than  formerly  to  the  principles  of 
illumination  and  to  photometry.  The  Johns  Hopkins  University- 
I.  E.  S.  lecture  course  on  illumination  did  much  to  arouse  ped- 
agogic interest  in  this  subject,  and  the  Society's  Committee  on 
Illuminating  Engineering  Education  is  engaged  in  an  effort  to 
apply  the  lecture  course  and  promote  the  further  specialization 
along  this  line  in  university  education. 

It  is  of  prime  importance  that  those  who  are  in  a  position  to 
influence  the  design  of  lighting  installations  should  possess  a 
knowledge  of  the  principles  of  good  illumination.  Educational 
work  in  this  connection  is  being  done  on  a  limited  scale  by  some 
of  the  larger  manufacturers  of  lamps  and  lighting  appliances 
through  lectures,  the  publication  of  bulletins,  and  through  free 
consulting  engineering  advice  to  customers.  A  number  of  the 
larger  lighting  companies  contribute  like  service  to  their  cus- 
tomers. The  Commercial  Section  of  the  National  Electric  Light 
Association  and  the  National  Commercial  Gas  Association  have 
done  commendable  work  in  the  publication  of  bulletins  on  cer- 
tain phases  of  lighting  work.  This  Society  has  in  course  of  pre- 
paration lectures  on  several  classes  of  lighting  the  manuscript 
of  which,  accompanied  by  lantern  slide  illustrations,  is  to  be  made 
available  for  presentation  wherever  required. 

Supplementing  these  efforts  is  the  dissemination  of  elementary 
knowledge  of  good  illumination  direct  to  the  public  through  the 
illumination  primer  prepared  by  this  Society.  Through  the 
efforts  of  the  Society  and  the  co-operation  of  other  organizations 
this  has  been  issued  throughout  the  country  in  quantities  which 
aggregate  about  250,000. 

Knowledge  is  power.  If  the  simple  elementary  truths  in  re- 
gard to  lighting  could  be  imparted  to  the  public,  lighting  practise 
the  country  over  would  be  advanced  tremenduously.  Through 
the  several  agencies  just  mentioned  it  is  believed  that  great 
strides  are  being  made  in  this  direction. 

COMMERCIAL  ORGANIZATIONS  IN  THE  LIGHTING  FIELD. 

In  this  country  the  very  general  enterprise  and  progressiveness 
of  the  leading  manufacturers  of  lamps  and  lighting  appliances  is 


676  TRANSACTIONS   I.   E.    S. — PART   II 

an  asset  upon  which  we  are  to  be  congratulated.  A  number  of 
such  organizations  maintain  laboratories  for  research  work  as 
well  as  engineering  and  educational  departments.  Thus  these 
manufacturing  organizations  contribute  to  the  advance  of  the 
lighting  art  in  three  ways.  First,  by  developments  made  pos- 
sible through  research  and  invention;  second,  by  the  incorpora- 
tion in  the  design  of  their  product  of  the  information  made  avail- 
able through  research  in  their  own  laboratories  and  elsewhere; 
third,  by  the  general  dissemination  of  the  facts  which  govern 
good  practise. 

The  progressive  attitude  of  lighting  companies,  particularly 
those  in  the  larger  cities  is  likewise  commendable  and  is  an  im- 
portant factor  in  lighting  improvement.  Through  the  training 
of  solicitors  in  lighting  fundamentals  and  the  maintenance  of  il- 
luminating engineering  departments,  as  well  as  through  adher- 
ence to  the  free  lamp  renewal  policy,  and  the  promotion  of  the 
use  of  the  most  efficient  lamps,  the  contributions  of  such  or- 
ganizations is  large  and  important. 

Another  agency  not  to  be  ignored  is  the  papers  and  discussions 
on  illumination  which  are  now  included  very  generally  in  the  pro- 
grams of  electric  and  gas  associations,  and  to  some  extent  in  the 
programs  of  professional  bodies.  Still  another  agency  is  the 
technical  journals,  which  now  very  generally  feature  illumination 
discussions,  thereby  doing  much  to  promote  interest  and  advance 
knowledge  in  illumination  affairs. 

THE   ILLUMINATING  ENGINEERING  SOCIETY. 

In  his  inaugural  address  the  writer  endeavored  to  review  the 
growth  of  the  society  and  to  state  its  functions  as  these  are 
now  recognized.  At  first,  scientific  discussions  predominated  be- 
cause knowledge  must  precede  application,  and  knowledge  of  il- 
lumination principles  was  meager.  The  tendency  during  recent 
years  has  been  to  supplement  the  scientific  work  with  practical 
discussions  of  lighting  practise  and  with  efforts  along  educational 
lines. 

The  only  organization  in  the  country  which  deals  exclusively 
with  light,  this  Society  now  commands  a  fair  measure  of  respect 
from  the  older  national  technical  societies.  It  is  beginning  to 
realize  returns  upon  its  large  expenditure  of  time  and  effort  in 


THE    STATUS   OF   THE   LIGHTING    ART  677 

co-operative  work  with  other  technical  societies.  It  is  undoubt- 
edly one  of  the  principal  influences  which  make  for  improvement 
in  the  lighting  art. 

THE  ATTITUDE  OF  RELATED  PROFESSIONS  AND  INDUSTRIES. 

That  the  efforts  of  the  Illuminating  Engineering  Society  to 
improve  lighting  practise  are  not  lacking  in  support  by  other  or- 
ganizations has  already  been  made  apparent.  It  has  been  indicated 
that  large  and  progressive  corporations  are  devoting  attention  to 
the  securing  of  good  illumination  as  a  part  of  good  management 
and  also  in  connection  with  welfare  work  for  employees.  Light- 
ing companies,  both  gas  and  electric,  are  showing  revived  appre- 
ciation of  the  importance  of  the  lighting  aspect  of  their  business. 
While  in  both  industries  the  lighting  load  is  decreasing  in  pro- 
portion to  the  total  load,  yet  it  is  recognized  that  through  its 
lighting  service  the  company  comes  into  contact  with  the  greatest 
numbers  of  its  public,  and  excellence  of  service  in  this  connection 
is  its  best  possible  advertisement. 

As  an  illustration  of  the  present  view  of  lighting  companies,  it 
may  be  said  that  in  the  1913  report  of  the  Lamp  Committee  of 
the  Association  of  Edison  Illuminating  Companies,  considerable 
space  was  devoted  to  the  importance  of  promoting  good  illumina- 
tion as  a  means  of  cultivating  good  public  opinion  and  conserving 
revenues. 

Perhaps  the  most  interesting  and  promising  features  of  the 
replies  received  in  connection  with  the  lighting  survey  which  is 
described  herein  is  the  very  evident  interest  manifested  by  oph- 
thalmologists throughout  the  country  in  the  subject  of  illumina- 
tion. There  is  undoubtedly  reason  to  believe  that  the  society  will 
enjoy  a  larger  measure  of  co-operation  from  ophthalmologists  in 
the  immediate  future  than  has  been  accorded  in  the  past,  and 
this  promises  well  for  improvement  in  home  conditions  of  light- 
ing. 

It  must  be  said  that  the  co-operation  of  architects  in  the  devel- 
opment of  lighting  practise  is  in  'need  of  further  cultivation. 
Attempts  during  the  past  year  to  promote  further  co-operative  ef- 
fort with  organizations  of  architects  have  come  to  naught,  although 
encouragement  was  derived  from  the  cordial  attitude  met  in  all 
cases.     It  is  hoped  that  a  start  may  be  made  during  the  coming 


678  TRANSACTIONS    I.    E.    S. — PART    II 

year  which  will  result  in  securing  recognition  of  mutuality  of 
interest  and  a  greater  measure  of  that  co-operative  effort  which 
is  so  necessary  to  the  improvement  in  lighting  practise  in  build- 
ings in  which  the  architect  requires  consistent  lighting  treatment. 

PHOTOMETRY. 

To  the  development  of  a  science,  proper  measurement  of  the 
quantities  involved  is  essential.  In  the  growth  of  knowledge  of 
illuminating  engineering,  photometry  has  played  an  important 
part.  At  the  present  time  the  status  of  photometry  is  definitely 
established,  and  very  generally  recognized.  In  the  manufacture 
of  illuminants  photometry  is  resorted  to  in  rating  and  efficiency 
adjustments.  In  taking  the  candle-power  of  gas  it  finds  one  of 
its  most  general  applications.  In  spite  of  the  increasing  strength 
of  the  movement  to  rate  gas  upon  calorific  rather  than  photo- 
metric value,  the  candle-power  basis  still  obtains  very  generally. 
In  the  production  of  incandescent  electric  lamps,  photometry 
has  found  one  of  its  widest  applications.  The  standardization  of 
manufacturing  methods,  however,  is  resulting  in  the  abandonment 
of  photometry  for  the  rating  of  individual  Mazda  lamps,  though 
•the  photometry  of  samples  from  each  batch  manufactured  still 
continues  to  be  a  regular  part  of  the  manufacturing  procedure. 
In  the  manufacture  of  other  illuminants,  photometry  is  a  regular 
part  of  the  engineering  work.  In  the  study  of  illumination,  pho- 
tometry is  practised  very  generally  by  illuminating  engineers. 
Accuracies  obtained  in  practise  in  the  various  classes  of  work 
are  of  the  order  indicated  in  the  following  table: 

Typical  Accuracies  Obtaining  in  Photometry  of  Light  Sources 

Which  Involves  no  Large  Color  Differences  and  no 

Serious  Variations  in  Light  Intensity. 

Per  cent. 

In  taking  the  candle-power  of  gas 2 

In  routine  rating  of  carbon  and  Gem  filament  lamps ±4 

In  routine  commercial  testing  of  incandescent  electric 

lamps  and  incandescent  gas  lamps —   ±2 

In  precision  photometry ±}4 

In  heterochromatic  photometry  lower  orders  of  accuracy  are 
encountered,  depending  upon  the  extent  of  the  color  difference 
and  the  methods  employed  in  determinations  of  intensities. 
Where  material  variations  in  intensities  are  involved,  further  in- 


THE   STATUS  OF   THE  LIGHTING  ART  679 

accuracies  may  result  due  to  personal  pecularities  of  observers 
in  recording  observations  under  such  conditions.  Here  again 
the  accuracy  which  is  obtained  depends  largely  upon  the  extent 
of  the  unsteadiness  of  the  light  source  and  the  methods  of  test 
employed. 

A  matter  of  immediate  importance  to  photometricians  is  the 
adoption  of  a  standard  for  use  in  heterochromatic  photometry. 
In  view  of  the  great  variety  of  light  sources  which  are  available, 
and  their  differing  color  characteristics,  it  is  important  that  stand- 
ards be  adopted  for  working  purposes,  even  though  such  stand- 
ards cannot  be  relied  upon  for  ultimate  accuracy.  In  the  adop- 
tion of  such  standards  it  is  believed  that  the  entire  laboratory 
resources  of  the  country  should  be  utilized.  This  is  one  of  the 
next  important  steps  to  be  taken  in  photometry. 

conclusion. 

The  foregoing  is  a  very  inadequate  review  of  the  condition  in 
the  field  which  this  Society  seeks  to  cultivate.  Circumstances 
beyond  control  interfered  with  the  writer's  intention  to  make  this 
survey  as  comprehensive  and  accurate  as  possible.  Instead  of 
presenting  a  complete  and  satisfactory  statement  which  might 
be  placed  upon  file  as  a  record  of  the  status  of  the  lighting  art, 
it  becomes  necessary  to  offer  this  review  as  a  first  step  toward 
the  preparation  of  such  a  record.  However,  discussion  of  the 
subject  with  members  of  this  Society,  and  experience  in  prepar- 
ing this  review,  have  brought  conviction  that  a  continuation  of 
this  effort  to  compile  an  adequate  compendium  of  the  knowledge 
of  the  art  and  of  prevailing  practise  would  result  in  much  good, 
and  would  repay  the  large  effort  which  would  be  required.  It  is 
therefore  without  apology  for  the  obvious  inadequacy  of  this 
presentation,  but  rather  with  the  expressed  hope  that  others  may 
consider  the  advisability  of  undertaking  a  more  thorough  and 
authoritative  treatment  that  this  survey  of  lighting  conditions  is 
presented. 

Our  knowledge  of  the  principles  of  good  illumination,  though 
lacking  in  many  essentials,  is  still  considerable.  The  advances 
made  since  the  organization  of  this  Society  in  1906  are  gratify- 
ing. The  condition  for  further  development  of  the  underlying 
principles  is  most  hopeful,  in  view  of  the  investigation  and 
9 


68o  TRANSACTIONS   I.    E.    S. — PART   II 

research  which  are  being  carried  on  by  a  goodly  number  of  our 
members.  Some  of  this  research  in  particular,  is  organized  upon 
such  a  plan  as  to  warrant  a  most  optimistic  view  of  the  probable 
extension  of  knowledge  in  this  field  in  the  next  few  years. 

The  standard  product  of  manufacturers  is  improved  rapidly 
in  accordance  with  additions  to  knowledge  of  lighting  principles. 
Such  improvements  in  the  case  of  lamps  or  appliances  which  have 
limited  life  and  must  be  replaced  periodically,  find  their  way  grad- 
ually into  most  installations.  In  the  case  of  appliances  which  do 
not  have  to  be  replaced  periodically,  such  as  fixtures,  lighting 
glassware,  etc.,  the  improvements  are  not  applied  so  generally  in 
lighting  practise.  It  is  perhaps  a  misfortune  that  such  appli- 
ances do  not  have  a  limited  life.  If  fixtures,  reflecting  glassware, 
etc.,  would  automatically  disintegrate  after  a  reasonable  period 
of  service,  the  commercial  incentive  offered  to  manufacturers  to 
improve  lighting  equipments  would  be  even  greater  than  it  is 
to-day,  and  the  public,  would  benefit  more  generally  as  a  result 
of  the  improvement  which  is  effected  in  more  recent  designs.  As 
it  is,  new  installations  benefit,  but  existing  installations  in  large 
part  obtain  little  or  no  advantage  from  the  more  recent  develop- 
ments. 

Perhaps  the  biggest  problem  to  be  solved  by  the  Society  in  its 
effort  to  secure  the  general  improvement  of  illumination  in  all 
classes  of  installations  is  that  of  displacing  antiquated  lighting 
appliances.  It  would  appear  that  we  must  look  to  those  who 
have  possible  commercial  advantage  to  derive  for  the  display  of 
enterprise  which  is  essential  to  the  general  application  of  better 
knowledge  of  illumination  principles  to  general  lighting  practise. 
These  are  the  manufacturers  of  the  equipments,  and  the  lighting 
companies  who  supply  the  service.  Co-operation  between  these 
two  classes  alone  can  bring  the  rapid  improvement  in  the  lighting 
art  which  conditions  demand  and  which  it  is  the  avowed  object 
of  this  Society  to  promote.  Improvement  in  artificial  lighting 
which  involves  a  substantial  increase  in  its  cost,  does  not  weigh 
as  heavily  in  the  scale  of  expenditures  as  would  a  corresponding 
increase  in  many  other  items  of  expense. 

Given  a  condition  in  which  it  is  apparent  that  the  public  and  a 
number  of  important  commercial  interests  will  benefit  alike  by 
improvement  in  lighting  practise,  and  a  total  cost  of  lighting 


THE   STATUS  OF   THE   LIGHTING   ART  68l 


which  is  insignificant  in  comparison  with  the  benefits  conferred 
by  the  lighting,  and  a  cost  of  improvement  in  such  lighting  which 
is  small  in  comparison  with  the  advantages  realized  through  such 
improvements,  it  would  appear  that  there  is  no  insurmountable 
obstacle  in  the  way  of  attaining  the  desired  result  of  generally 
improved  lighting.  All  that  is  necessary  is  a  co-ordinating  in- 
fluence and  the  necessary  conviction  as  to  the  results  to  be  ob- 
tained.    Herein  lies  this  Society's  opportunity. 

There  is  reason  to  be  dissatisfied  with  progress.  Practise  lags 
inexcusably  behind  knowledge  and  ideals.  It  will  be  conceded 
that  good  illumination  is  greatly  to  be  desired,  the  same  conclu- 
sion being  reached  whether  the  viewpoint  is  dominantly  commer- 
cial, esthetic  or  humanitarian.  In  improved  lighting,  benefits 
accrue  to  the  public,  to  consulting  engineers,  to  lighting  com- 
panies, to  manufacturers  of  lamps,  to  manufacturers  of  auxil- 
iaries, and  to  manufacturers  of  fixtures.  To  improve  illumina- 
tion conditions,  the  public  must  be  aroused  to  an  appreciation  of 
the  advantages  ever  associated  with  such  improvement.  This  can 
be  done  if  all  the  interests  just  mentioned  can  combine  and  co- 
operate to  support  the  propaganda  of  this  Society.  The  enthu- 
siasm and  conviction  which  would  necessarily  characterize  such 
a  movement  would  constitute  an  irresistible  force  which  would 
be  certain  to  accomplish  the  purpose  of  awakening  public  interest 
and  bringing  conviction  of  the  benefits  which  would  be  sure  to 
follow.  How  are  we  to  enlist  the  support  and  co-operation  of 
the  commercial  interests  whose  participation  in  such  a  forward 
movement  is  essential  to  its  success?  This  is  the  practical  ques- 
tion which  this  Society  must  answer.  Already  the  beginning  has 
been  made,  and  the  Society  now  enjoys  a  measure  of  sympathy 
and  support  from  the  commercial  interests  with  which  it  has 
never  before  been  favored.  However,  it  is  necessary  to  con- 
vince the  lighting  industry  of  the  potentialities  of  the  situation 
before  any  really  effective  campaign  can  be  waged  on  a  large 
scale.  What  measures  can  be  adopted  for  bringing  this  about  ? 
Two  are  suggested  herewith. 

First,  it  is  recommended  that  the  Society  supplement  its  splen- 
did work  of  research,  discussion  and  education,  by  devoting 
special  considerations  to  ways  and  means  of  adopting  its  work  to 


682  TRANSACTIONS    I.    E.    S. PART    II 

meet  the  requirements  of  the  commercial  interests  to  which  we 
must  look  for  extension  of  improved  lighting  practise.  A  com- 
mittee on  commercial  application  would  doubtless  do  much  to 
further  this  cause. 

Second,  the  members  of  this  Society  individually  may  accom- 
plish results  which  in  the  aggregate  will  be  far-reaching  in  their 
effect  in  popularizing  good  illumination.  Lighting  which  is  in- 
adequate, inartistic,  unhygienic  or  inefficient  is  often  observed  in 
the  homes  of  the  members  of  this  Society.  Comment  brings 
some  such  reply  as  that  "He  that  makes  shoes  goes  barefoot  him- 
self." It  is  submitted  that  this  state  of  affairs  is  a  reproach  under 
which  no  member  of  this  Society  should  remain.  It  behooves 
each  of  us  to  combine  practise  with  precept.  Surely  there  is  no 
member  of  this  Society  who  does  not  feel  assured  that  by  light- 
ing his  home  properly,  he  will  contribute  to  the  happiness  and 
welfare  of  the  members  of  his  family,  perhaps  more  largely  than 
is  possible  with  a  similar  expenditure  in  any  other  way. 
Then  what  excuse  can  there  be  for  the  member  of  this  Society, 
who  presumably  is  informed  in  these  matters,  who  recognizes  the 
importance  and  the  possibilities  of  good  lighting,  and  yet 
permits  the  retention  in  his  home  or  office  of  improper  lighting 
equipments  ? 

Furthermore,  it  may  be  assumed  that  most  of  the  members  of 
this  Society  have  more  or  less  commercial  interest  in  the  ex- 
tension of  good  lighting.  On  narrower  and  more  sordid  grounds, 
is  it  not  incumbent  upon  such  members  to  promote  good  lighting 
practise  by  example  ? 

Should  we  not  attend  to  the  installation  in  our  homes  of  mod- 
ern lighting  equipment  which  is  in  reasonable  conformity  with 
the  latest  tenets  of  illuminating  engineering?  Also  should  we 
not  employ  such  equipment  freely,  to  the  end  that  in  our  homes 
we  may  at  all  proper  times  exemplify  the  faith  which  is  within 
us?  How  great  an  influence  upon  the  community  would  be  ex- 
erted if  every  man  who  is  interested  in  the  furtherance  of  our 
propaganda  would  so  illuminate  his  home  that  it  would  demon- 
strate to  all  who  observe,  the  advantages  of  good  lighting!  It 
ought  to  be  possible  to  say  of  the  members  of  this  Society — "By 
their  lighting  ye  shall  know  them." 


ANNUAL  REPORT  OF  THE  GENERAL  SECRETARY  683 

ANNUAL  REPORT  OF  THE  GENERAL  SECRETARY 
FOR  THE  FISCAL  YEAR  ENDING  SEPTEMBER  30,  1913. 


The  past  fiscal  year  of  the  Society  has  been  one  of  general 
expansion.  The  educational  and  co-operative  work  started  in 
the  previous  year  has  been  continued  and  extended,  while  new 
lines  for  developing  the  Society  and  enlarging  its  sphere  and  in- 
fluence have  been  followed  with  gratifying  results.  Perhaps  in 
no  other  year  has  such  a  general  concerted  effort  been  made  to 
place  the  Society  on  a  broad  and  firm  basis  for  future  service. 
Some  of  the  results  achieved  are  briefly  recounted  in  the  follow- 
ing report. 

1.      FINANCES. 

The  financial  status  (see  auditor's  report  on  another  page  of 
this  issue)  of  the  Society  has  been  improved  somewhat  by  the 
acquisition  of  sustaining  members.  Although  the  expenses  have 
increased  because  the  work  of  the  Society  has  been  conducted 
along  more  extensive  and  useful  lines,  the  revenue  has  been 
slightly  in  excess  of  the  expenses. 

Under  the  following  captions,  Income  and  Expense  statistics 
are  given  to  show  the  extent  of  the  various  sources  of  income 
and  expense.  The  per  cent,  figures  which  are  fairly  indicative 
may  be  used  for  comparisons  with  other  years;  but  the  amounts 
are  not  to  be  compared  with  similar  figures  without  considering 
the  fact  that  the  past  fiscal  year  was  of  only  nine  months  dura- 
tion. 

Income. — About  63  per  cent,  of  the  total  income  was  derived 
from  members'  dues,  and  8  per  cent,  from  sustaining  members. 
Both  sources  netted  $5,897.08,  or  71  per  cent,  of  the  total  rev- 
enue. Only  about  28  per  cent,  or  $2,359.41,  therefore,  was  ob- 
tained from  other  sources,  including  $1,097.14  or  approximately 
13  per  cent,  from  advertising. 

Expenses. — An  analysis  of  the  expenses  shows  that  $3,768.84 
or  46  per  cent,  was  expended  for  general  office  salaries,  rent, 
supplies,  etc.  The  next  largest  item  was  that  of  the  Transac- 
tions, $1,844.72,  or  22  per  cent,  of  the  total  expenses. 

The  latter  sum  is  relatively  25  to  30  per  cent,  higher  than  the 
cost  of  the  Transactions  for  a  similar  period  of  last  year,  on 


684  TRANSACTIONS    I.    E.    S. PART    II 

account  of  the  new  form  and  style  which  was  adopted  at  the 
beginning  of  the  present  year. 

Exclusive  of  the  latter  two  items,  $2,562.09,  or  31  per  cent, 
was  required  for  general  Society  expenses. 

Another  interesting  fact  is  that  the  total  revenue  from  the 
dues  of  members  and  sustaining  members,  amounted  to  72  per 
cent,  of  the  total  expenses.  Without  the  revenue  from  sustain- 
ing members,  the  income  from  dues  was  only  63  per  cent,  of  the 
total  expenses. 

The  surplus  of  $1,453.94  as  of  January  1,  1913,  has  been  in- 
creased to  $1,865.40  as  of  September  30,  191 3. 

The  total  assets  of  the  Society  are  shown  to  be  $6,838.18, 
against  which  there  are  liabilities  of  $4,972.78. 

II.       MEMBERSHIP. 

A  gain  was  also  made  in  the  membership.  One  hundred  and 
forty-eight  applications  and  re-instatements  were  received ;  while 
the  defections  totaled  86 — a  net  gain  of  62  members. 

In  addition  to  the  foregoing  increase  21  sustaining  members, 
whose  names  are  listed  below,  were  elected.  This  class  of  mem- 
bership was  created  by  a  change  in  the  constitution  which  became 
effective  at  the  beginning  of  the  present  fiscal  year.  It  is  gratify- 
ing to  note  in  passing  the  number  of  organizations  which  have 
so  promptly  applied  for  membership. 

Name  Date  of  election 

Electrical  Testing  Laboratories 3/T4/T3 

Holophane  Works  of  General  Electric  Company 3/14/13 

The  Edison  Electric  Illuminating  Company  of  Boston  • .  3/14/13 

The  New  York  Edison  Company 3/14/13 

The  Philadelphia  Electric  Company 3/14/13 

The  Edison  Electric  Illuminating  Company  of  Brooklyn  4/11/13 

Commonwealth  Edison  Company 4/11/13 

Macbeth- Evans  Glass  Company 4/1 1/13 

Westinghouse  Lamp  Company 4/11/13 

Boston  Consolidated  Gas  Company 4/1 1/13 

National  Electric  Lamp  Association 4/1 1/13 

Benjamin  Electric  Manufacturing  Company 4/11/13 

United  Electric  Light  &  Power  Company 5/  9/13 

Welsbach  Company 6/13/13 

Consolidated  Gas,  Electric  Light  and  Power  Company 

of  Baltimore 6/13/13 

Alexalite  Company 9i22>il$ 


ANNUAL  REPORT  OF   THE  GENERAL   SECRETARY  685 

Name  Date  of  election 

Cooper-Hewitt  Electric  Company 9/23/13 

Jefferson  Glass  Company 9 23/13 

Little  Rock  Railway  &  Electric  Company 923/13 

Pittsburgh  Lamp,  Brass  &  Glass  Company 9/23/13 

The  Leeds  and  Northrup  Company 9/23/13 

The  sources  of  membership  gains  and  losses  are  shown  in 
the  tabulation  given  below : 

Sections  Un- 

affiliated 


Phila-  New  Pitts-  New  and 

delphia       York       Chicago       burgh    England   foreign 
Total  number  of  members 

at    beginning    of    year, 

January  1,   1913 338  399  209  159  95  135 

Total  applications  received      13  59  33  14  2  27 

Defections     (resignations 

and  deceased  members)        9  29  15  13  n  9 

Total   number  at   end   of 

year 342  429  227  160  86  153 

It  will  be  observed  that  the  only  loss  of  membership  was  in 
the  New  England  Section.     From  territories  without  the  juris- 
diction of  sections,  27  applications  and  9  defections  have  been 
received,  making  a  net  gain  of  18  members. 
Through  death  the  following  members  were  lost: 

Carrigan,  Howard  F. 

1538  First  National  Bank  Building,  Chicago,  Ills. 

Douglass,  David 

Eau  Claire  Gas  Light  Company,  Eau  Claire,  Wis. 

Schniewind,  Dr.  F., 

6  Church  Street,  New  York  City. 

A  geographical  distribution  of  the  Society's  members  is  shown 
in  the  following  tabulation  : 

Illuminating  Engineering  Society  Members. 
United  Stales. 

Sustaining 
Members      members 

Alabama 2 

Arkansas 1 

California 19 

Colorado 7 

Connecticut  13 

District  of  Columbia 17 

Florida 3 


686 


TRANSACTIONS    I.    E.    S. — PART    II 


United  States — {continued). 

Members 

Georgia 7 

Illinois 129 

Indiana 20 

Iowa 10 

Kansas 6 

Kentucky 2 

Louisiana  4 

Maine 4 

Maryland   21 

Massachusetts 76 

Michigan    • 21 

Minnesota    9 

Missouri  17 

Nebraska    4 

New  Hampshire 4 

New  Jersey 91 

New  York 339 

North  Carolina 2 

Ohio 92 

Oklahoma 2 

Oregon 7 

Pennsylvania 34S 

Rhode  Island ...    5 

South  Carolina 2 

South  Dakota 2 

Tennessee 1 

Texas 4 

Utah 4 

Vermont 1 

Virginia 6 

Washington 5 

West  Virginia 3 

Wisconsin    26 

i,335 
Members  not  in  the  United  States. 

Canada 21 

England  17 

France 3 

Germany 6 

Hawaii    1 

Japan 1 

Mexico 1 

Panama •  •  1 

Philippines •" 2 

South  America 8 

Spain 1 

62 
Total  Members 1.335  +  62  =  1,397 


Sustaining 
members 


ANNUAL  REPORT  OF   THE  GENERAL  SECRETARY  687 

III.      SECTIONS. 

The  following  table  of  comparisons  gives  a  bird's  eye  view 
of  the  work  and  progress  of  the  several  sections  of  the  Society : 

Phila-  New  Pitts-  New 

delphia  York  Chicago  burgh        England 

Total  members  Sept.  30, 

1913 342  429              227               160              86 

Net  changes  during  year  +4  +3°  +  *8  +1  — 9 
Number  of  meetings  held  65564 
Number  of  papers  printed 

in  Transactions 2*  5                1*              3*             3 

Average     attendance     at 

meetings 121  144                68                37               35 

Total  expenses,  9  months  $130.88  $245.03       $217.04       $149.73       $53-78 
*  One  paper  pending  publication. 

Each  one  of  the  several  sections  has  had  one  or  more  joint 
meetings  with  other  Societies.  These  meetings  have  for  the  most 
part  been  devoted  to  popular  and  elementary  discussions  of 
lighting  subjects.  Consequently  the  contributions  to  the  Trans- 
actions have  been  small  compared  with  previous  years.  These 
meetings,  however,  which  were  intended  to  be  of  an  educational 
character,  have  strengthened  the  prestige  of  the  Society  and  won  a 
great  deal  of  respect  for  the  science  and  art  of  illumination  from 
many  societies  which  have  hitherto  had  only  a  vague  idea  of 
what  is  illuminating  engineering.  In  general  the  success  and  re- 
sults of  section  activities  have  been  very  gratifying. 

LOCAL  SECRETARIES. 

To  extend  the  work  and  influence  of  the  Society  to  those  cities 
which  do  not  have  sections,  several  local  secretaries  have  been 
appointed.     The  local  secretaries  are : 

State  City  Name  and  Address 

California Los  Angeles R.  H.  Manahan,  City  Electrician 

California San  Francisco. .'.    .  F.    Emerson     Hoar,     Railroad 

Commission  of  State  of  Cali- 
fornia, S32  Market  Street. 

Colorado Denver G.  E.Williamson,  Denver  Gas  & 

Electric  Light  Co. 

Georgia Atlanta William  Rawson  Collier, Georgia 

Railway  and  Light  Co. 


688  TRANSACTIONS    I.    E.    S. PART    II 

State                                               City  Name  and  Address 

Minnesota Minneapolis G.  D.  Shepardson,  University  of 

Minnesota,       Minneapolis, 

Minn. 
Minnesota St.  Paul A.    L.     Abbott,     Northwestern 

Electric  Equipment  Co. 
Washington Seattle Fred.  A.  Osborn,  University  of 

Washington. 

It  is  expected  that  these  representatives  will  endeavor  to  pro- 
mote occasional  meetings  under  the  joint  auspices  of  the  Illumi- 
nating Engineering  Society  and  local  organizations  with  a  view 
to  fostering  interest  in  lighting  matters.  Such  activities  should 
lead  eventually  to  the  organization  of  sections  in  the  given  cities 
or  territories. 

TRANSACTIONS. 

Eighteen  papers  on  various  phases  of  illuminating  engineering 
have  been  published  in  the  Transactions  during  the  past  year. 
From  the  list  of  titles  given  below  it  will  be  noted  that  these 
papers  include  a  wide  range  of  subjects.  Taken  together  they  are 
fairly  indicative  of  the  field  covered  by  the  science  and  art  of 
illumination. 

Department  Store  Lighting. 

Influence  of  Colored  Surroundings  on  the  Color  of  the  Useful  Light. 

The  Theory  of  Mercury-Vapor  Apparatus. 

Tests  for  the  Efficiency  of  the  Eye  under  different  Systems  of  Illumina- 
tion, and  a  Preliminary  Study  of  the  Causes  of  Discomfort. 

Street  Lighting  with  Ornamental  Luminous  Arc  Lamps. 

Some  Phases  of  the  Illumination  of  Interiors. 

Illumination  and  Eye  Strain. 

Home  Illumination. 

A  Photometer  Screen  for  use  in  Tests  of  Street  Illumination. 

The  Flame  Carbon  Arc  Lamp. 

The  Illumination  of  Motion  Picture  Projectors. 

Street  Lighting  of  Greater  New  York. 

The  Illumination  of  Passenger  Cars. 

Some  Home  Experiments  in  Illumination  from  Large  Area  Light 
Sources. 

Gas  Lighting  in  an  Exhibition  Hall. 

Metal  Reflectors  for  Industrial  Lighting. 

Vision  as  Influenced  by  the  Brightness  of  Surroundings. 

A  Practical  Solution  of  the  Problem  of  Heterochromatic  Photometry. 


ANNUAL  REPORT  OF   THE  GENERAL   SECRETARY  689 

Nearly  all  the  papers  and  many  of  their  attending  discussions 
have  laid  more  or  less  stress  upon  the  hygienic  aspects  of  light 
and  illumination.  Greatest  interest  has  been  manifest  in  those 
papers  dealing  with  the  commercial  application  of  the  scientific 
principles  to  lighting  practise,  say  in  the  design  or  re-design  of 
installations. 

Not  any  of  the  papers  presented  at  the  191 3  convention  are 
included  in  the  foregoing  list. 

Beginning  with  the  first  of  the  year  the  Transactions  was 
published  in  a  new  form.  The  general  make-up  of  the  publica- 
tion has  been  improved.  Paper  free  from  glare  has  been  used 
throughout.  The  new  style  has  increased  the  cost  of  publi- 
cation between  25  and  30  per  cent. 

committies. 

Some  twenty-five  permanent  and  temporary  committees  have 
conducted  the  work  of  the  society  with  rather  unusual  activity. 
A  brief  record  of  the  success  of  their  work  is  given  in  the  follow- 
ing paragraphs. 

1913  Convention  Committee. — The  191 3  convention — thanks 
to  the  good  work  of  the  committee — surpassed  in  general 
excellence  any  previous  convention.  The  attendance  of  out-of- 
town  delegates  was  considerably  larger  than  that  of  any  previous 
year.  An  excellent  program  of  papers,  well  balanced  with  com- 
mercial and  scientific  subjects,  lively  discussions,  and  a  generous 
complement  of  amusement  combined  to  make  the  convention  an 
unusual  success. 

Committee  on  the  Glore  from  Reflecting  Surfaces. — This  com- 
mittee followed  up  the  work  which  it  started  in  1912.  It  has 
confined  its  efforts  chiefly  to  giving  publicity  to  the  evils  arising 
from  the  glare  of  reflected  surfaces,  especially  from  glazed 
paper,  and  to  the  collection  of  data  and  information  pertaining 
to  the  use  of  unglazed  paper.  An  eight-page  leaflet  entitled 
"Glare"  was  published  and  circulated.  The  leaflet  consisted  of 
glazed  and  unglazed  paper  and  was  designed  primarily  for  the 
purpose  of  emphasizing  the  evils  of  glare  from  paper.  The  com- 
mittee has  communicated  with  a  number  of  leading  publishers  of 


69O  TRANSACTIONS    I.    E.    S. — PART    II 

books  and  periodicals.  Several  publishers  have  already  adopted 
unglazed  paper  through  the  efforts  of  the  committee.  In  the 
opinion  of  the  committee  glazed  paper  will  be  eliminated  from 
general  use  as  soon  as  this  change  can  be  effected  without  an 
unreasonable  increase  in  cost.  With  an  increased  demand  for 
unglazed  paper,  it  is  very  likely  that  an  entirely  satisfactory  paper 
will  be  forthcoming. 

Committee  on  Research. — The  committee  has  drafted  some  ex- 
cellent plans  for  its  procedure.  If  these  are  pursued  even  to  a 
small  degree  in  the  future,  noteworthy  results  must  accrue. 
Briefly  the  committee  has  planed  to  make  its  services  (1)  co- 
operative and  advisory,  and  (2)  initiatory.  The  committee  favors 
co-operation  with  several  committees  of  the  Society  and  sister 
societies,  with  departments  of  universities,  research  laboratories 
and  individuals,  when  matters  of  research  are  concerned.  The 
committee  proposes  (a)  to  suggest  research  problems  to  parties 
desirous  and  competent  to  do  valuable  work,  (b)  to  compile  a 
list  of  subjects  in  which  research  seems  to  be  advisable,  and 
(c)  to  arrange  a  bibliography  on  the  various  phases  of  illumina- 
tion. While  the  fulfillment  of  these  functions  and  plans  has  not 
progressed  sufficiently  to  warrant  definite  achievements,  results 
may  be  expected  in  the  future. 

Committee  on  Collegiate  Education. — The  work  of  this  com- 
mittee has  been  conducted  in  accordance  with  the  following  plan ; 
(1)  to  determine  which  colleges  and  universities  are  giving  any 
attention  to  illuminating  engineering;  (2)  to  find  out  exactly 
the  character  and  extent  of  such  work  as  is  being  done;  (3)  to 
call  attention  of  college  authorities  to  the  importance  of  greater 
and  adequate  instruction  in  all  matters  pertaining  to  the  use  of 
light,  with  a  view  to  (a)  bringing  about  the  introduction  of 
courses  in  illumination,  and  (b)  looking  to  the  extension  and 
amplification  of  present  courses.  To  ascertain  the  status  of  il- 
luminating engineering  in  the  curricula  of  colleges,  the  committee 
has  sent  a  letter  to  a  number  of  college  presidents,  asking  for 
information,  in  regard  to  courses  in  illumination  in  their  insti- 
tutions. A  sufficent  number  of  returns  have  not  yet  been  re- 
ceived from  these  inquiries  to  warrant  a  statement  at  this  time.. 


ANNUAL  REPORT  OF  THE  GENERAL  SECRETARY  691 

Committee  on  Reciprocal  Relations  with  other  Societies. — 
The  excellent  work  of  this  committee,  which  was  appointed 
for  the  first  time  in  1912,  has  been  continued  during  the  present 
year.  The  committee  has  been  in  touch  with  some  twenty  or 
thirty  societies  and  has  arranged  joint  meetings  with  several  of 
them,  notably  the  American  Gas  Institute,  The  National  Com- 
mercial Gas  Association,  The  National  Electric  Light  Associa- 
tion, The  American  Academy  of  Medicine,  the  School  Hygiene 
Congress  and  the  American  Medical  Association.  In  acquaint- 
ing other  organizations  with  the  objects  and  province  of  the 
Illuminating  Engineering  Society  this  committee  has  rendered 
a  service  of  great  value. 

Committee  on  Sustaining  Membership. — Through  the  efforts 
of  this  committee  the  companies  whose  names  appear  on  a  pre- 
vious page  all  became  sustaining  members  within  a  period  of  five 
months.  The  revenue  from  sustaining  membership  has  increased 
the  funds  of  the  society  so  that  the  work  on  a  more  enlarged  and 
efficient  scale  might  be  started. 

Committee  on  Nomenclature  and  Standards. — The  committee 
has  held  several  meetings,  and  has  been  in  touch  with  the  stan- 
dards committees  of  several  American  and  foreign  societies, 
which  are  interested  in  the  science  and  art  of  illumination.  It 
has  formulated  plans  for  aiding  the  formation  of  an  international 
photometric  commission.  It  has  also  proposed  a  number  of  re- 
vised definitions  of  standards  and  of  terminology,  which  have 
been  transmitted  to  the  International  Photometric  Commission 
and  interested  societies. 

Committee  of  Factory  Lighting  Legislation. — Appointed  this 
year  for  the  first  time  to  consider  matters  of  legislation  where  il- 
lumination is  concerned,  this  committee  has  undertaken  and  ren- 
dered valuable  service  in  a  field  which  heretofore  has  received 
scant  and  insufficient  attention.  Upon  invitation,  its  members 
attended  a  meeting  of  the  New  York  State  Factory  Commission 
and  made  a  number  of  recommendations  relating  to  the  lighting 
of  factories  and  workrooms,  with  special  reference  to  adequacy 
and  quality  of  illumination.     A  bill  incorporating  these  recom- 


692  TRANSACTIONS    I.    E.    S. PART    II 

mendations  was  submitted  to  the  Legislature  and  became  a  law 
in  New  York  State  on  October  1,  1913.  The  committee  was 
also  represented  at  a  conference  of  the  Heights  of  Buildings 
Committee  of  the  Board  of  Estimate  and  Apportionment  of  the 
City  of  New  York,  at  which  the  illumination  of  interiors  by  day- 
light and  by  artificial  light  was  discussed. 

Section  Development  Committee. — During  the  year  the  com- 
mittee completed  work  on  the  preparation  of  a  guide  on  section 
management  which  had  been  started  in  the  previous  year.  The 
guide  constitutes  a  convenient  arrangement  of  notes  and  sug- 
gestions which  tells  in  a  general  way  how  the  affairs  of  section 
may  be  advantageously  conducted.  It  is  believed  that  it  will  go 
a  long  way  toward  promoting  co-operation,  co-ordination  and 
efficiency  in  the  work  of  the  society.  Heretofore  new  boards  of 
managers  have  had  to  undertake  the  management  of  sections 
with  a  very  indefinite  knowledge  of  the  duties  and  responsibilities 
to  be  discharged.  Now  any  board,  though  it  be  entirely  new, 
may  in  an  hour  or  two  become  familiar  with  what  has  been 
found  to  be  good  procedure  and  what  work  was  done  by  the  pre- 
vious board.  Hereafter  it  is  likely  that  the  preliminary  plans 
and  details  incidental  to  the  beginning  of  each  season  will  be 
arranged  with  greater  dispatch.  Copies  of  the  guide  have  been 
issued  to  the  members  of  all  the  section  boards  as  well  as  to 
a  number  of  officers  of  the  society. 

Committee  on  Finance. — The  Committee  on  Finance  has  exer- 
cised general  supervision  over  the  affairs  of  the  Society.  It  has 
approved  all  appropriations  and  disbursements. 

Committee  on  Illumination  Primer. — A  few  slight  revisions 
were  made  in  the  first  edition  of  the  primer  which  was  published 
last  fall.  Of  this  revised  edition  approximately  a  quarter  of  a 
million  copies  have  been  printed.  Over  220,000  copies  of  the 
latter  edition  were  sold  to  lighting  companies  throughout  the 
country  for  distribution  to  their  customers.  The  primer  has  been 
widely  advertised  and  commended  in  trade,  and  other  pournals 
both  in  this  country  and  abroad.  No  work  of  the  Society  has 
met  with  greater  approval  on  the  part  of  the  public  and  the  light- 
ing industry  in  general.    Certainly  no  publication  of  its  kind  has 


ANNUAL  REPORT  OF   THE  GENERAL   SECRETARY  693 

done  more  to  spread  the  gospel  of  good  illumination,  particularly 
in  the  way  of  creating  a  desire  for  better  standards  of  illumina- 
tion. 

Committee  on  Papers. — The  Committee  on  Papers  has  passed 
upon  all  papers  submitted  to  the  Society  within  the  past  year. 
It  has  earned  the  distinction  of  having  arranged  for  the  191 3 
convention  a  program  which  has  been  pronounced  the  best  bal- 
anced set  of  papers  ever  presented  before  a  meeting  of  the 
Society. 

Committee  on  Editing  and  Publication. — While  practically  all 
the  publication  work  of  the  Society  has  been  done  in  the  general 
offices,  the  Editing  Committee  has  served  in  an  advisory  capacity 
and  passed  upon  from  time  to  time  questions  involving  the  edi- 
torial policy  of  the  Society.  A  guide  for  authors,  setting  forth 
the  editorial  requirements  and  style  of  papers  has  been  prepared 
for  distribution.  The  committee  believes  that  this  pamphlet  will 
promote  a  degree  of  desirable  uniformity  in  all  papers  and  at  the 
same  time  obviate  considerable  expense  in  publication. 

Committee  on  Advertising. — In  accordance  with  instructions 
from  the  Council,  the  committee  has  made  no  special  attempt  to  in- 
crease the  amount  of  advertising  contracted  for  during  the  year 
1912.  The  net  income  derived  from  advertising,  $1,097.14,  was 
relatively,  a  little  higher  than  that  for  a  corresponding  period  of 
1912. 

Committee  on  Membership. — No  active  general  campaign  for 
increasing  membership  has  been  conducted.  Through  the  chair- 
man of  the  various  section  Membership  Committees,  a  conserva- 
tive effort  was  made  to  bring  into  the  Society  those  members 
who  are  most  likely  to  contribute  to  and  profit  by  its  work.  Dur- 
the  year  the  total  membership  was  increased  by  62  members. 
One  hundred  and  forty-eight  applications  and  reinstatements 
were  received  largely  as  a  result  of  the  committee's  efforts. 

Committee  on  Progress. — The  work  which  this  committee  did 
is  indicated  by  the  excellent  and  comprehensive  report  of  progress 
in  the  field  of  illuminating  engineering  during  the  past  year, 
which  was  submitted  at  the  convention.  The  report*  sets  a  high 
standard  of  comparison  for  all  future  progress  reports. 

*  Trans.  I.  E.  S.,  Vol.  VIII,  No.  7  (October,  1913),  p.  323- 


694  TRANSACTIONS    I.    E.    S. PART    II 

Committee  on  Joint  Session  with  the  American  Gas  Institute. — 
This  committee  was  appointed  solely  for  the  purpose  of  arrang- 
ing an  illuminating  engineering  session  at  the  annual  meeting  of 
the  American  Gas  Institute  which  was  held  in  Richmond  in 
October,  1913.  For  this  session  the  committee  arranged  for  a 
number  of  papers  and  talks  on  various  phases  of  illuminating 
engineering  and  the  work  of  the  Society  itself. 

Committee  on  Joint  Meeting  with  the  International  Congress 
of  School  Hygiene. — A  session  at  which  several  papers  on  the 
hygienic  aspects  of  light  and  illumination  were  presented  was 
arranged  by  the  committee.  The  Congress  manifested  a  great 
deal  of  interest  in  this  particular  session. 

Committee  on  1915. — This  committee  was  appointed  for  the 
purpose  of  considering  the  advisability  of  holding  the  191 5 
convention  of  the  Society  in  San  Francisco  during  the  Panama 
Pacific  Exposition.  No  definite  recommendations  were  made  be- 
cause the  committee  believed  the  time  of  that  convention  was 
too  far  off. 

Committee  on  Popular  Lectures. — The  Committee  on  Popular 
Lectures  was  appointed  at  the  beginning  of  the  year  to  make 
available  popular  lectures  on  good  lighting  practice.  With  such 
lectures  it  is  hoped  that  it  will  be  possible  to  create  on  the  part 
of  the  public  a  better  appreciation  of  good  lighting  and  ultimately 
raise  the  standards  of  illumination.  While  it  is  planned  to  cover 
other  fields  of  lighting,  the  first  efforts  will  be  confined  to  resi- 
dence, store,  industrial  lighting  and  office  lighting.  Sub-com- 
mittees have  been  appointed — one  for  each  of  these  four  classes 
of  lighting,  the  members  of  which  are  experts  familiar  with  the 
best  practice.  In  order  that  lectures  may  be  useful,  without 
evincing  commercial  prejudice,  a  special  endeavor  has  been  made 
to  balance  the  committees  with  reference  to  the  interests  repre- 
sented that  they  may  be  free  from  any  commercial  bias.  It  is 
proposed  that  each  of  these  sub-committees  shall  prepare  one  or 
more  lectures  on  the  subjects  assigned.  Photographs  from  which 
lantern  slides  may  be  made  will  be  collected  to  illustrate  by 
example  what  is  considered  good  lighting  practise,  and  on  ihe 


ANNUAL   REPORT  OF   THE   GENERAL   SECRETARY  695 

other  hand  to  emphasize  the  evils  of  poor  illumination.  A  careful 
selection  of  the  committees  has  just  been  completed  and  it  is 
planned  to  have  available  within  the  next  month  or  two  at  the 
general  offices  of  the  Society,  a  series  of  lectures  which  may  be 
loaned  to  societies  and  associations,  lighting  companies,  both  gas 
and  electric,  and  other  organizations  which  are  interested  in 
illumination. 

To  the  Committee  of  Tellers  which  counted  with  patience  and 
care  the  ballots  of  the  annual  election,  and  to  the  Committee  on 
the  Formation  of  a  Lake  Erie  Section,  which  considered  and  re- 
ported the  possibilities  of  such  a  section,  special  thanks  is  due. 
The  latter  proposal  while  encouraging  was  left  for  future  de- 
velopment. 

Delegates. — During  the  year  a  number  of  representatives  have 
been  appointed,  upon  invitation,  to  boards  and  committees  of 
other  organizations,  (International  Gas  Congress,  Committee  on 
Organization  International  Electrical  Congress,  American  Asso- 
ciation for  the  Conservation  of  Vision,  United  States  National 
Committee  of  the  International  Commission  on  Illumination)  in 
order  that  the  Society  might  keep  in  touch  with  work  with  which 
it  is  concerned.  The  services  of  these  representatives  deserves 
the  thanks  of  the  Society. 

general. 

Here,  at  the  end  of  another  year,  there  is  cause  to  look  back 
with  grateful  appreciation  upon  praiseworthy  efforts  of  the 
officers,  committeemen,  section  boards,  and  many  other  members 
who  have  given  unstintingly  of  their  time  and  thought  to  the 
work  and  objects  which  the  Society  seeks  to  promote.  The 
measure  of  success  attained  must  be  ascribed  entirely  to  their 
work. 

Worthy  of  record,  also,  is  the  encouraging  co-operation  which 
has  been  obtained  from  sister  organizations  in  promoting  what 
has  been  aptly  called  the  gospel  of  better  illumination.  There 
have  been  numerous  evidences  of  increasing  and  greater  respect 
for  illuminating  engineering  or  the  science  and  art  of  illumina- 
tion as  many  prefer  to  call  it.  Bonds  of  common  interest  be- 
tween the  I.  E.  S.  and  many  societies  have  been  established  and 
strengthened. 


696  TRANSACTIONS    I.    E.    S. — PART    II 

Further,  the  initial  activities  of  the  German  Illuminating  En- 
gineering Society,  and  the  promise  of  organization  of  a  French 
society  before  long,  have  attracted  world-wide  attention.  In  a 
word,  illuminating  engineering  is  gradually  coming  into  its  own. 
Let  those  who  wish  predict  when  illuminating  engineering  will 
take  its  place  among  the  firmly  established  professions.  For  the 
present  it  is  gratifying  enough  to  observe  the  rapid  advancement 
in  that  direction. 

Some  reason  there  might  be  for  referring  in  this  report  to  the 
generally  consistent  progress,  the  notable  developments,  and  the 
numerous  examples  of  scientifically  designed  installations,  in  the 
lighting  field  during  the  past  year  if  all  these  were  not  already 
chronicled  elsewhere. 

Respectfully  submitted, 

Joseph  D.  Israel, 

General  Secretary. 


TP 
700 
133 
v.  8 


Illuminating  engineering 


^frifili    '   ~ 

Engineering 

PLEASE  DO  NOT  REMOVE 
CARDS  OR  SLIPS  FROM  THIS  POCKET 

UNIVERSITY  OF  TORONTO  LIBRARY 


ENGIN 


STORAGE